Repository object · research-note
Candidate Dossier
Accepted research note in the public catalog.
- Media type
application/json- Object ID
em:research-note:sha256:ed2271f5334ac51f552a491c70e8394c115332b7e7a77be11b4a375d8a361645- Content digest
1ae06b54fe6c6ce1803836bbf2ecaf3e652bed2c6878b7e095c01a1c689ab87b
Also filed under
Source content
{
"assertions": [
{
"actor": {
"id": "V2-SOL-01",
"kind": "agent"
},
"asserted_at": "2026-08-23T18:33:06Z",
"id": "em:dossier-assertion:sha256:57d53dc8cda085d335aba2fc377aca36f22666902606696640f5f322aaea90ca",
"key": "assertion-v2-sol-01-a73232acf8",
"lineage_key": "lineage-v2-sol-01-a73232acf8",
"proposition_key": "prop-report-citation-observation",
"span_keys": [
"span-audit-report-1"
],
"stance": "asserts",
"visibility": "public"
},
{
"actor": {
"id": "V2-SOL-02",
"kind": "agent"
},
"asserted_at": "2026-08-23T18:33:06Z",
"id": "em:dossier-assertion:sha256:e5c14d00f5c8fc8cfb0c07f07e8a3e0cc8dda56b754c220ba22db0c0d16173e8",
"key": "assertion-v2-sol-02-82e3cd76cd",
"lineage_key": "lineage-v2-sol-02-82e3cd76cd",
"proposition_key": "prop-report-citation-observation",
"span_keys": [
"span-audit-report-2"
],
"stance": "asserts",
"visibility": "public"
},
{
"actor": {
"id": "V2-SOL-03",
"kind": "agent"
},
"asserted_at": "2026-08-23T18:33:06Z",
"id": "em:dossier-assertion:sha256:db0a5239838bbbfda29d9b946e93a251bd5a56481387812769eed9c4082f54cd",
"key": "assertion-v2-sol-03-a54afd37f7",
"lineage_key": "lineage-v2-sol-03-a54afd37f7",
"proposition_key": "prop-report-citation-observation",
"span_keys": [
"span-audit-report-3"
],
"stance": "asserts",
"visibility": "public"
},
{
"actor": {
"id": "V2-SOL-04",
"kind": "agent"
},
"asserted_at": "2026-08-23T18:33:06Z",
"id": "em:dossier-assertion:sha256:31e38ccb01249a7a7a86bb45d335b0983384d07cb30bcd750b718d55ccb1fa71",
"key": "assertion-v2-sol-04-7c709ab690",
"lineage_key": "lineage-v2-sol-04-7c709ab690",
"proposition_key": "prop-report-citation-observation",
"span_keys": [
"span-audit-report-4"
],
"stance": "asserts",
"visibility": "public"
},
{
"actor": {
"id": "V2-TERRA-01",
"kind": "agent"
},
"asserted_at": "2026-08-23T18:33:06Z",
"id": "em:dossier-assertion:sha256:1d3196577c9061708f720a4e86375f1fbff114fba447334a91e86d1f18574316",
"key": "assertion-v2-terra-01-962446cd82",
"lineage_key": "lineage-v2-terra-01-962446cd82",
"proposition_key": "prop-report-citation-observation",
"span_keys": [
"span-audit-report-5"
],
"stance": "asserts",
"visibility": "public"
},
{
"actor": {
"id": "V2-TERRA-02",
"kind": "agent"
},
"asserted_at": "2026-08-23T18:33:06Z",
"id": "em:dossier-assertion:sha256:13cecca59936403573783ec38a18bab84796f4e88b82c5af44b25f5535138a07",
"key": "assertion-v2-terra-02-7e359a7894",
"lineage_key": "lineage-v2-terra-02-7e359a7894",
"proposition_key": "prop-report-citation-observation",
"span_keys": [
"span-audit-report-6"
],
"stance": "asserts",
"visibility": "public"
},
{
"actor": {
"id": "V2-TERRA-03",
"kind": "agent"
},
"asserted_at": "2026-08-23T18:33:06Z",
"id": "em:dossier-assertion:sha256:49dec33e7ffdcfd2708fb5a7fa15818f9360e2624989d8241cd71cc1e0da17ee",
"key": "assertion-v2-terra-03-c20170eb34",
"lineage_key": "lineage-v2-terra-03-c20170eb34",
"proposition_key": "prop-report-citation-observation",
"span_keys": [
"span-audit-report-7"
],
"stance": "asserts",
"visibility": "public"
},
{
"actor": {
"id": "V2-TERRA-04",
"kind": "agent"
},
"asserted_at": "2026-08-23T18:33:06Z",
"id": "em:dossier-assertion:sha256:ffaaf74efb9518caab6f2f082b8f5bae8a23bf985a4ed348e2ffe3a25004e289",
"key": "assertion-v2-terra-04-d04d23915b",
"lineage_key": "lineage-v2-terra-04-d04d23915b",
"proposition_key": "prop-report-citation-observation",
"span_keys": [
"span-audit-report-8"
],
"stance": "asserts",
"visibility": "public"
},
{
"actor": {
"id": "work:keplinger-dermatology-audit+work:keplinger-supplement",
"kind": "collective"
},
"asserted_at": "2026-08-23T18:33:06Z",
"id": "em:dossier-assertion:sha256:dbab0bf3479337837fb7d341657082207174d8a39ff61d224eb2d5ac228fe8be",
"key": "assertion-keplinger-metadata-versus-support-73dfc141c5",
"lineage_key": "lineage-keplinger-metadata-versus-support-73dfc141c5",
"proposition_key": "prop-keplinger-metadata-versus-support-73dfc141c5",
"span_keys": [
"span-092745c7db6407d4d52927642741cde229fad05a5347bb18-7b04227f8f",
"span-39a9931f05ea356fe904ed15768fa08c66440b1fcb20f98e-191c24bbb2",
"span-46c392bc942cd88d525d74298e6bc6ebd9aeaec533ddfb7b-774f5b996e",
"span-8c8d9a6dc003bf0298542b85df3f6f08db631e68e89f606f-92731379f2",
"span-94ad583514aa7e04b8d5f7b8f9cdda6fdffe8ca6712f033d-63c1ba24eb",
"span-b38a2af5982b4c85bc205d2f533a23ed3a8f40a49b651a5f-27395a6291",
"span-b6a6ba48a2c4e352c8565b671a15a90400739f2aeefda8a6-e8232daf27",
"span-b903dc2284e1b4d80bb6ef948d79bbdbc7d7447e871281b8-6ef6234316",
"span-f05b3189dc3fafd45b5bde643a91fd0616832928215a37a2-47f72e77e4"
],
"stance": "asserts",
"visibility": "public"
},
{
"actor": {
"id": "work:deepresearch-bench-paper+work:deepresearch-bench-repository",
"kind": "collective"
},
"asserted_at": "2026-08-23T18:33:06Z",
"id": "em:dossier-assertion:sha256:7a83bdb7d484e15702c9b201fd817128505788f16fbe8738b028fab0ee45a0f8",
"key": "assertion-deepresearch-bench-fact-ab2a1c0fd8",
"lineage_key": "lineage-deepresearch-bench-fact-ab2a1c0fd8",
"proposition_key": "prop-deepresearch-bench-fact-ab2a1c0fd8",
"span_keys": [
"span-015bb68c904bf225e84c82b1acffb3cc38a4a087d7b0f79d-4c454092b1",
"span-0d4f357f2a5565bdaea01a1e537650958d71216868d1a2b0-9b0af35d76",
"span-13cb0ddb087b82988b7445a9cfb02562f9c044ef4f244e8b-87c3b42c68",
"span-196c54d6212fbc0c054dbf8e8467f2388c1f788bb93f4b1a-d460e927f6",
"span-279db725380c133044944dccb379b75a396242219aa30437-569fc9e2fd",
"span-2f311bfb451c6208101cb7583af37c338914fcdb10f5c75b-90e7764ce2",
"span-7d7331942f9d0520e18e96206e7718e5187788be4d48cef9-dfdbc1d019",
"span-92eb7cf19745e24190dda84c77d590941134d09f06805c0f-88afe43377",
"span-9fe42cb48702d8d96ebbbc0309692214ece6ec495c447c06-80da541506",
"span-b0a39c963cdfdc5f9852cfc9f7e33e05d31232bd1d899885-adf9545222",
"span-b83996e3b5fdf878e04d6d41d0e7a1eebee9fdd9bbccf2ae-820a3eab86",
"span-c27aaca02bda06ed764be53351158fc862af9d4a556d3e82-104e7cda52",
"span-e4a9f30effc031388ed5c9993dc6f071903bec5e3f1998db-5d95269b76",
"span-e5bcf0241a1aeae7d9a92755b1bcf0adeeea93a8922f22f8-77a5ff4cc0",
"span-eaf37bef7b8dc35f3e00b85d3e58ad872dc0ea604a062c8c-fa09300f85",
"span-ef637b38f36195bea09205d532e8744daf3e257f93d97144-26007673a3",
"span-f98f1274859ba98cc538e3be6d11d345a5382cfcb9516aa1-d60b765c48"
],
"stance": "asserts",
"visibility": "public"
},
{
"actor": {
"id": "work:deeptrace",
"kind": "collective"
},
"asserted_at": "2026-08-23T18:33:06Z",
"id": "em:dossier-assertion:sha256:b041cab9ed6954b90c0160b9679e6b36a6048e41734b20e431c3875574dbaee9",
"key": "assertion-deeptrace-support-variation-de1dfdaff3",
"lineage_key": "lineage-deeptrace-support-variation-de1dfdaff3",
"proposition_key": "prop-deeptrace-support-variation-de1dfdaff3",
"span_keys": [
"span-35927851edd2e7e5e0f60d98498940f88304ba99bf1f85a0-2fe586436b",
"span-8cce3d9ee1e1674ba1b93321dd9b99a55d834ca980451f8a-d2fceafe11",
"span-9cb1700c08368feb234207045354cbf559d6ed80f2b0bae8-f898722eb5",
"span-9dd6eb46529e77a62748093528e52685fbd92d46a756aef7-96d75d9e2b",
"span-a08468cb62f2feb21d9cbb95175ce00f54592bd83573cc91-03f65cfa45",
"span-a15ab02e9fcb23df02fb03d28967d4220ddf65ad380d2230-ddf85fa359",
"span-aa1534ce2e177153ab85e512deb8d064d14ec1eae16d8032-2299096a09",
"span-e2c9f6220843a04665f5d1cd142ac15f04211ace5c579fa6-c5e0476df0",
"span-supplement-deeptrace-gemini-prose-40-3-ed29752a6f",
"span-supplement-deeptrace-gemini-table-50-3-7eb7bd254b"
],
"stance": "asserts",
"visibility": "public"
},
{
"actor": {
"id": "work:cited-not-verified",
"kind": "collective"
},
"asserted_at": "2026-08-23T18:33:06Z",
"id": "em:dossier-assertion:sha256:6fa46782b3e3128aca1e97f75cba4846be90f31c236e01ed2968ef92ac7c1378",
"key": "assertion-cnv-link-relevance-support-gap-81b86b6d92",
"lineage_key": "lineage-cnv-link-relevance-support-gap-81b86b6d92",
"proposition_key": "prop-cnv-link-relevance-support-gap-81b86b6d92",
"span_keys": [
"span-732c03e69c6e142a492275cee5d1e2d474c75422c9b84aaa-f7675998cc",
"span-74f6acafbb4c54d556a572460bf1b45bacd1e293f4aa16cc-5ee15c82c4",
"span-912c931063109f84ea88aea34891dd5e0f9147d2176af258-4d566ca076",
"span-bb16d5f06abe8641d9ea6694e549aa15e240dd88946aa47a-5946b6e6b7",
"span-cd21c450c0d33df20a2b1a539d5725347db1b5bab7d1e986-2e3e0c3ac7",
"span-deada8bacb987a47f2a42e52fbfeb5c1beb3d250c0be79e2-0e91f5cacc"
],
"stance": "asserts",
"visibility": "public"
},
{
"actor": {
"id": "work:cited-not-verified",
"kind": "collective"
},
"asserted_at": "2026-08-23T18:33:06Z",
"id": "em:dossier-assertion:sha256:366913c92b23900577dce66033da69d3c1d7373df2855a30421e3d0e7e2127b3",
"key": "assertion-cnv-search-depth-ablation-dad05fa3f9",
"lineage_key": "lineage-cnv-search-depth-ablation-dad05fa3f9",
"proposition_key": "prop-cnv-search-depth-ablation-dad05fa3f9",
"span_keys": [
"span-417182b4c15ac550469e435a02aab1e91b395f72f186312d-e9e2c4a70c",
"span-6e961498ed0f81a950f775956587f741e0eb9f11640fd3cb-afa5b59854",
"span-c3dfc98566c27561eb7267972e00e3e1ba6a399a3ff5d002-37f40d3260",
"span-d638c8b21f0e4a0f840880b6ffe0217ce1623758cc09fdcc-8585f6d39b"
],
"stance": "asserts",
"visibility": "public"
},
{
"actor": {
"id": "work:url-health",
"kind": "collective"
},
"asserted_at": "2026-08-23T18:33:06Z",
"id": "em:dossier-assertion:sha256:c1b99004fca0476c7c180f7ea829071b94278f2a2b5754678d316d098fe19c6b",
"key": "assertion-url-health-resolution-9fcccc534a",
"lineage_key": "lineage-url-health-resolution-9fcccc534a",
"proposition_key": "prop-url-health-resolution-9fcccc534a",
"span_keys": [
"span-55de6a6ca9414a252592e9f77544a9259533dd713ce1f12f-0826216ffc",
"span-a18e778fac3d49e81167f05e09fbc361e91af8c0c01b0fec-ca1b18e206",
"span-bb62928b5e0cb4e373b2f6bfeb579d8564d05da5588f8b15-771310f4ef",
"span-supplement-url-health-drbench-precollected-outputs-44923dff13"
],
"stance": "asserts",
"visibility": "public"
},
{
"actor": {
"id": "work:reportbench",
"kind": "collective"
},
"asserted_at": "2026-08-23T18:33:06Z",
"id": "em:dossier-assertion:sha256:2d0c2401cc47abfb1b09b0cd2facf39f6b0e5e9da4c2ff28f3d61679df670365",
"key": "assertion-reportbench-match-rate-46ddf0a876",
"lineage_key": "lineage-reportbench-match-rate-46ddf0a876",
"proposition_key": "prop-reportbench-match-rate-46ddf0a876",
"span_keys": [
"span-1db12b68cfeb4c5f5a96fa9cd2e2bd46e8dfb4f096f73154-cea47f0c65",
"span-3015a109e7898cb720e426e9a57c526277b648c75f1fb4b1-745d84ff44",
"span-6c1d327e58ef001533ea1c11010e974436bbeafbde138efc-f6efd8d376",
"span-7a6046a756d6ec0315d330cf0817d062ea48498bdc9c90b3-c3a1e80029",
"span-9c7a4c186cc85f185aa293a7c5a46c08f9dbbc1a9e63c449-feff07a6e2",
"span-ba6b8a8b7d9121ebb05d594176ceed3d1e9dbc637876783a-d85147b6b0",
"span-ddf4320a6d302c13d28b6225eb7e79ac15583349e1b449d4-f89bb46179",
"span-e223640cb0ed12c87ca1d2406f3276f30a3b8d2017dd4dc1-3cfecd76bf"
],
"stance": "asserts",
"visibility": "public"
},
{
"actor": {
"id": "EM-0026 independent semantic review",
"kind": "service"
},
"asserted_at": "2026-08-23T18:33:06Z",
"id": "em:dossier-assertion:sha256:61b5c6d72986bb26c4afd8ace2dcfceef65bd39fcffcc4f59a2d967742077bc5",
"key": "assertion-citation-verifier-calibration-5b40082faa",
"lineage_key": "lineage-citation-verifier-calibration-5b40082faa",
"proposition_key": "prop-citation-verifier-calibration-da47a19a65",
"span_keys": [
"span-audit-pending-1"
],
"stance": "questions",
"visibility": "public"
},
{
"actor": {
"id": "EM-0026 independent semantic review",
"kind": "service"
},
"asserted_at": "2026-08-23T18:33:06Z",
"id": "em:dossier-assertion:sha256:b3e23c0d5edad672eea86d63069aee9997e11f713c0a4229ffba5d0122df3d5b",
"key": "assertion-liveresearchbench-e1-e2-e3-5fb5d671f7",
"lineage_key": "lineage-liveresearchbench-e1-e2-e3-5fb5d671f7",
"proposition_key": "prop-liveresearchbench-e1-e2-e3-07e84c4837",
"span_keys": [
"span-audit-pending-2"
],
"stance": "questions",
"visibility": "public"
},
{
"actor": {
"id": "EM-0026 independent semantic review",
"kind": "service"
},
"asserted_at": "2026-08-23T18:33:06Z",
"id": "em:dossier-assertion:sha256:dcdeb5f34f1dc2781afc1f747bcb5c7c58f49c1562ba3cbbbdc79502fe68f34b",
"key": "assertion-researcherbench-faithfulness-groundedness-c84dcaee54",
"lineage_key": "lineage-researcherbench-faithfulness-groundedness-c84dcaee54",
"proposition_key": "prop-researcherbench-faithfulness-groundedness-0deb7adf49",
"span_keys": [
"span-audit-pending-3"
],
"stance": "questions",
"visibility": "public"
},
{
"actor": {
"id": "EM-0026 independent semantic review",
"kind": "service"
},
"asserted_at": "2026-08-23T18:33:06Z",
"id": "em:dossier-assertion:sha256:80207035190acce7bb460635b2d2919fbbe4bd3ce40220f9bf300d4ccfaa666c",
"key": "assertion-url-health-correction-loop-eb8a6384c6",
"lineage_key": "lineage-url-health-correction-loop-eb8a6384c6",
"proposition_key": "prop-url-health-correction-loop-f63084450b",
"span_keys": [
"span-audit-pending-4"
],
"stance": "questions",
"visibility": "public"
},
{
"actor": {
"id": "EM-0026 deterministic audit",
"kind": "instrument"
},
"asserted_at": "2026-08-23T18:33:06Z",
"id": "em:dossier-assertion:sha256:94eab68a46a3ec78e4b5684a929282c7e203119cb8ef14d94a1056b6ac647fac",
"key": "assertion-derived-counts",
"lineage_key": "lineage-em0026-audit",
"proposition_key": "prop-relation-derived-counts",
"span_keys": [
"span-audit-counts"
],
"stance": "asserts",
"visibility": "public"
},
{
"actor": {
"id": "EM-0026 deterministic audit",
"kind": "instrument"
},
"asserted_at": "2026-08-23T18:33:06Z",
"id": "em:dossier-assertion:sha256:fd24064e6f0574ed8c3db473d6ac2462c2c27bddb8f16ec056e283cfe72ba4b5",
"key": "assertion-unresolved-citations",
"lineage_key": "lineage-em0026-audit",
"proposition_key": "prop-unresolved-citations-no-credit",
"span_keys": [
"span-audit-unresolved"
],
"stance": "asserts",
"visibility": "public"
},
{
"actor": {
"id": "EM-0026 deterministic audit",
"kind": "instrument"
},
"asserted_at": "2026-08-23T18:33:06Z",
"id": "em:dossier-assertion:sha256:05c7b2842e6715a58bc22867a9bf66049d9bf8866aee45c9ddefacd559097315",
"key": "assertion-inaccessible-carriers",
"lineage_key": "lineage-em0026-audit",
"proposition_key": "prop-inaccessible-carriers-no-credit",
"span_keys": [
"span-audit-inaccessible"
],
"stance": "asserts",
"visibility": "public"
},
{
"actor": {
"id": "EM-0026 deterministic audit",
"kind": "instrument"
},
"asserted_at": "2026-08-23T18:33:06Z",
"id": "em:dossier-assertion:sha256:f8961ef802ffa486da8b28014bcc8079aa22f74b44e32fc92241efa9f0b63177",
"key": "assertion-unsupported-claims",
"lineage_key": "lineage-em0026-audit",
"proposition_key": "prop-unsupported-claims-no-credit",
"span_keys": [
"span-audit-unsupported"
],
"stance": "asserts",
"visibility": "public"
},
{
"actor": {
"id": "EM-0026 deterministic audit",
"kind": "instrument"
},
"asserted_at": "2026-08-23T18:33:06Z",
"id": "em:dossier-assertion:sha256:d62b7612f3f212d3769089cd03969e49ff5a914ee8067149e329b3ae5d8a12e5",
"key": "assertion-independently-rejected-claims",
"lineage_key": "lineage-em0026-audit",
"proposition_key": "prop-independently-rejected-claims-no-credit",
"span_keys": [
"span-audit-rejected"
],
"stance": "asserts",
"visibility": "public"
}
],
"claim_families": [
{
"assertion_keys": [
"assertion-v2-sol-01-a73232acf8",
"assertion-v2-sol-02-82e3cd76cd",
"assertion-v2-sol-03-a54afd37f7",
"assertion-v2-sol-04-7c709ab690",
"assertion-v2-terra-01-962446cd82",
"assertion-v2-terra-02-7e359a7894",
"assertion-v2-terra-03-c20170eb34",
"assertion-v2-terra-04-d04d23915b",
"assertion-keplinger-metadata-versus-support-73dfc141c5",
"assertion-deepresearch-bench-fact-ab2a1c0fd8",
"assertion-deeptrace-support-variation-de1dfdaff3",
"assertion-cnv-link-relevance-support-gap-81b86b6d92",
"assertion-cnv-search-depth-ablation-dad05fa3f9",
"assertion-url-health-resolution-9fcccc534a",
"assertion-reportbench-match-rate-46ddf0a876",
"assertion-citation-verifier-calibration-5b40082faa",
"assertion-liveresearchbench-e1-e2-e3-5fb5d671f7",
"assertion-researcherbench-faithfulness-groundedness-c84dcaee54",
"assertion-url-health-correction-loop-eb8a6384c6",
"assertion-derived-counts",
"assertion-unresolved-citations",
"assertion-inaccessible-carriers",
"assertion-unsupported-claims",
"assertion-independently-rejected-claims"
],
"id": "em:dossier-claim-family:sha256:c997ee2fdd3f78ae937ef6267fb637272bcb8e0cfd64e75a1a877677623c9e0c",
"key": "family-agent-citation-lineage",
"proposition_keys": [
"prop-agreement-not-independent-warrant",
"prop-report-citation-observation",
"prop-unresolved-citations-no-credit",
"prop-unsupported-claims-no-credit",
"prop-inaccessible-carriers-no-credit",
"prop-independently-rejected-claims-no-credit",
"prop-relation-derived-counts",
"prop-keplinger-metadata-versus-support-73dfc141c5",
"prop-deepresearch-bench-fact-ab2a1c0fd8",
"prop-deeptrace-support-variation-de1dfdaff3",
"prop-cnv-link-relevance-support-gap-81b86b6d92",
"prop-cnv-search-depth-ablation-dad05fa3f9",
"prop-url-health-resolution-9fcccc534a",
"prop-url-health-correction-loop-f63084450b",
"prop-reportbench-match-rate-46ddf0a876",
"prop-researcherbench-faithfulness-groundedness-0deb7adf49",
"prop-liveresearchbench-e1-e2-e3-07e84c4837",
"prop-citation-verifier-calibration-da47a19a65"
],
"question": "When multiple research-agent reports cite overlapping material, what does their agreement add beyond the inspected source and warrant lineages?",
"relation_keys": [
"relation-v2-sol-01-e7797662c8",
"relation-v2-sol-02-8b059b3c99",
"relation-v2-sol-03-6b89558412",
"relation-v2-sol-04-5e52e77bc3",
"relation-v2-terra-01-d2701a39bc",
"relation-v2-terra-02-0ca185da71",
"relation-v2-terra-03-672f07dbef",
"relation-v2-terra-04-b8e0fb7c26",
"relation-keplinger-metadata-versus-support-c23e105c8b",
"relation-keplinger-metadata-versus-support-d9ee7c5a8a",
"relation-deepresearch-bench-fact-a3f61d7a3a",
"relation-deepresearch-bench-fact-2efd170c89",
"relation-deeptrace-support-variation-1a84ef7c6f",
"relation-deeptrace-support-variation-678e984a6a",
"relation-cnv-link-relevance-support-gap-165443d270",
"relation-cnv-link-relevance-support-gap-911219619e",
"relation-cnv-search-depth-ablation-9c01321092",
"relation-cnv-search-depth-ablation-d7aecf2d25",
"relation-url-health-resolution-0b26fb45c7",
"relation-url-health-resolution-b1cc744437",
"relation-reportbench-match-rate-6964e41a69",
"relation-reportbench-match-rate-4f30573b54",
"relation-citation-verifier-calibration-5b40082faa",
"relation-liveresearchbench-e1-e2-e3-5fb5d671f7",
"relation-researcherbench-faithfulness-groundedness-c84dcaee54",
"relation-url-health-correction-loop-eb8a6384c6",
"relation-assertion-derived-counts-234c8dbdaf",
"relation-assertion-unresolved-citations-35a17beebf",
"relation-assertion-inaccessible-carriers-a9d4494367",
"relation-assertion-unsupported-claims-30720e36be",
"relation-assertion-independently-rejected-claims-4af09a9196"
],
"title": "Report, citation, source, span, and warrant lineage",
"visibility": "public"
}
],
"dossier_id": "em:dossier:sha256:cbd7a14096a956f642f5c76046d3b49ed648fbe6bf24144c992404a01415af82",
"editions": [
{
"content": {
"canonical_url": "https://arxiv.org/html/2607.08700v1",
"excerpts": [
{
"license_treatment": "quote-minimal attributed spans",
"locator": "Section 1 contributions",
"match_status": "exact-normalized-match",
"occurrence_ids": [
"V2-SOL-03:s6:s6_span1"
],
"source_text_sha256": "0bd9a0bedc4b8ea9b08689109a9a6d763cb78d7513df2598874f147afa9ff9f7",
"span_root_id": "span:sha256:6885cf2524462022475b5829da081b2d35fcfe7463ca6ad0979e7c4fb1f0b54d",
"text": "624 attribution-citation pairs with gold labels for all 1,248 LLM-judged decisions, every one human-reviewed"
},
{
"license_treatment": "quote-minimal attributed spans",
"locator": "Abstract, line 16",
"match_status": "exact-normalized-match",
"occurrence_ids": [
"V2-SOL-04:S7:S7-SP2"
],
"source_text_sha256": "0bd9a0bedc4b8ea9b08689109a9a6d763cb78d7513df2598874f147afa9ff9f7",
"span_root_id": "span:sha256:794bdf7e0595799afbb0d55f3dc522077aed9615c59f1e3575707148cf444bcf",
"text": "378 of which were hard cases"
},
{
"license_treatment": "quote-minimal attributed spans",
"locator": "Abstract, line 16",
"match_status": "exact-normalized-match",
"occurrence_ids": [
"V2-SOL-04:S7:S7-SP1"
],
"source_text_sha256": "0bd9a0bedc4b8ea9b08689109a9a6d763cb78d7513df2598874f147afa9ff9f7",
"span_root_id": "span:sha256:e54ee1457854a9a5e46f7c7c2e66ac6a5a69b963f8ee3450f40052f138d5d0c1",
"text": "1,248 rubric decisions"
}
],
"format": "epistemedia-quote-minimal-edition-projection-v0.1",
"license": "CC BY 4.0",
"license_treatment": "quote-minimal attributed spans",
"readback_receipts": [
{
"captured_bytes": 43218,
"captured_sha256": "c3efca0e3c20c437e63b3a990c85b2392f0244d3dc8194312a8394923a16cbec",
"edition_id": "edition:citation-verifier-arxiv-v1",
"http_status": 200,
"media_type": "text/html; charset=utf-8",
"redirect_chain": [],
"requested_url": "https://arxiv.org/abs/2607.08700",
"resolved_url": "https://arxiv.org/abs/2607.08700",
"retrieval_status": "retrieved"
}
],
"review_supplements": [],
"source_text_file": "citation-verifier.txt"
},
"content_digest": "sha256:837372b8c39f6d1cf4459dd4969695c2c8bba340dfa1cfd9e1bd5e0015a1eec7",
"content_length": 1903,
"edition_label": "Quote-minimal projection of edition:citation-verifier-arxiv-v1",
"id": "em:dossier-edition:sha256:11f00fc6576322cec62671d5ff7dec3c0c9bed06c139e6f704054259b06241c4",
"key": "edition-citation-verifier-arxiv-v1-acfd4abab7",
"media_type": "application/json",
"retrieved_at": "2026-08-23T18:33:06Z",
"visibility": "public",
"work_key": "work-citation-verifier-benchmark-2d5e94336b"
},
{
"content": {
"canonical_url": "https://arxiv.org/html/2605.06635v1",
"excerpts": [
{
"license_treatment": "quote-minimal attributed spans",
"locator": "Section 4.3, line 212",
"match_status": "exact-normalized-match",
"occurrence_ids": [
"V2-SOL-04:S1:S1-SP3"
],
"source_text_sha256": "1dd04031e118646bdebec5e35d289cc625e758ff165a63ba3cbf0680ad12bb10",
"span_root_id": "span:sha256:417182b4c15ac550469e435a02aab1e91b395f72f186312de9d34d57fd1c6c98",
"text": "from 79% to 17% (62%)"
},
{
"license_treatment": "quote-minimal attributed spans",
"locator": "Limitations, lines 219-224",
"match_status": "ordered-fragment-match",
"occurrence_ids": [
"V2-TERRA-03:S1:S1-F"
],
"source_text_sha256": "1dd04031e118646bdebec5e35d289cc625e758ff165a63ba3cbf0680ad12bb10",
"span_root_id": "span:sha256:5cd9ba4ca72da10cf51c0beddc6ea2765c74721bdd526a67156f44754219f440",
"text": "the LLM-as-a-judge approach ... may retain biases inherent to the judge model"
},
{
"license_treatment": "quote-minimal attributed spans",
"locator": "Section 4.3, Tables 2–3",
"match_status": "exact-normalized-match",
"occurrence_ids": [
"V2-SOL-03:s5:s5_span2"
],
"source_text_sha256": "1dd04031e118646bdebec5e35d289cc625e758ff165a63ba3cbf0680ad12bb10",
"span_root_id": "span:sha256:6e961498ed0f81a950f775956587f741e0eb9f11640fd3cb40870585558d869d",
"text": "Fact Check accuracy drops approximately 42% on average from minimal (2 calls) to maximal search depth"
},
{
"license_treatment": "quote-minimal attributed spans",
"locator": "Section 3.3.1-3.3.3",
"match_status": "exact-normalized-match",
"occurrence_ids": [
"V2-TERRA-04:S5_cited_not_verified:S5_method"
],
"source_text_sha256": "1dd04031e118646bdebec5e35d289cc625e758ff165a63ba3cbf0680ad12bb10",
"span_root_id": "span:sha256:732c03e69c6e142a492275cee5d1e2d474c75422c9b84aaaf273100e937338ed",
"text": "\"Fact Check verifies whether specific factual claims are accurately supported by the source content.\""
},
{
"license_treatment": "quote-minimal attributed spans",
"locator": "Abstract, lines 47-48",
"match_status": "exact-normalized-match",
"occurrence_ids": [
"V2-SOL-04:S1:S1-SP1"
],
"source_text_sha256": "1dd04031e118646bdebec5e35d289cc625e758ff165a63ba3cbf0680ad12bb10",
"span_root_id": "span:sha256:74f6acafbb4c54d556a572460bf1b45bacd1e293f4aa16cc2d499174096b9c20",
"text": "yet achieve only 39–77% factual accuracy"
},
{
"license_treatment": "quote-minimal attributed spans",
"locator": "Abstract and Table 1",
"match_status": "exact-normalized-match",
"occurrence_ids": [
"V2-SOL-02:S7:S7_SPAN_1"
],
"source_text_sha256": "1dd04031e118646bdebec5e35d289cc625e758ff165a63ba3cbf0680ad12bb10",
"span_root_id": "span:sha256:912c931063109f84ea88aea34891dd5e0f9147d2176af258059246431e9abfb1",
"text": "link validity above 94% and relevance above 80%, yet achieve only 39–77% factual accuracy"
},
{
"license_treatment": "quote-minimal attributed spans",
"locator": "Abstract and Table 1",
"match_status": "exact-normalized-match",
"occurrence_ids": [
"V2-SOL-03:s5:s5_span1"
],
"source_text_sha256": "1dd04031e118646bdebec5e35d289cc625e758ff165a63ba3cbf0680ad12bb10",
"span_root_id": "span:sha256:bb16d5f06abe8641d9ea6694e549aa15e240dd88946aa47a2d4e30ddd3351a24",
"text": "maintain link validity above 94% and relevance above 80%, yet achieve only 39–77% factual accuracy"
},
{
"license_treatment": "quote-minimal attributed spans",
"locator": "page 5, Section 3.3.3",
"match_status": "exact-normalized-match",
"occurrence_ids": [
"V2-SOL-01:s4_cited_not_verified:sp4c"
],
"source_text_sha256": "1dd04031e118646bdebec5e35d289cc625e758ff165a63ba3cbf0680ad12bb10",
"span_root_id": "span:sha256:c3dfc98566c27561eb7267972e00e3e1ba6a399a3ff5d002dca2d1233836790c",
"text": "contradicted, absent, or uncertain"
},
{
"license_treatment": "quote-minimal attributed spans",
"locator": "Section 5 'Limitations'",
"match_status": "ordered-fragment-match",
"occurrence_ids": [
"V2-TERRA-04:S5_cited_not_verified:S5_limitations"
],
"source_text_sha256": "1dd04031e118646bdebec5e35d289cc625e758ff165a63ba3cbf0680ad12bb10",
"span_root_id": "span:sha256:c7dbd1d9e09980703ecbee8161594e4b426c1af63117b28662a7e98686107621",
"text": "\"LLM-as-a-judge ... may retain biases inherent to the judge model\""
},
{
"license_treatment": "quote-minimal attributed spans",
"locator": "Abstract, lines 45-49",
"match_status": "exact-normalized-match",
"occurrence_ids": [
"V2-TERRA-03:S1:S1-A"
],
"source_text_sha256": "1dd04031e118646bdebec5e35d289cc625e758ff165a63ba3cbf0680ad12bb10",
"span_root_id": "span:sha256:cd21c450c0d33df20a2b1a539d5725347db1b5bab7d1e9866c24790d845e5873",
"text": "Citations are evaluated along three dimensions. (1) Link Works verifies URL accessibility, (2) Relevant Content measures topical alignment, and (3) Fact Check validates factual accuracy against source content."
},
{
"license_treatment": "quote-minimal attributed spans",
"locator": "Section 4.3, Tables 2-3 and following paragraph",
"match_status": "exact-normalized-match",
"occurrence_ids": [
"V2-TERRA-04:S5_cited_not_verified:S5_depth"
],
"source_text_sha256": "1dd04031e118646bdebec5e35d289cc625e758ff165a63ba3cbf0680ad12bb10",
"span_root_id": "span:sha256:d638c8b21f0e4a0f840880b6ffe0217ce1623758cc09fdccc32cacf2db3dd958",
"text": "\"Fact Check accuracy drops approximately 42% on average\""
},
{
"license_treatment": "quote-minimal attributed spans",
"locator": "Section 4.4, line 217",
"match_status": "exact-normalized-match",
"occurrence_ids": [
"V2-SOL-04:S1:S1-SP2"
],
"source_text_sha256": "1dd04031e118646bdebec5e35d289cc625e758ff165a63ba3cbf0680ad12bb10",
"span_root_id": "span:sha256:deada8bacb987a47f2a42e52fbfeb5c1beb3d250c0be79e2f9c5cdd6e16b4bc7",
"text": "only 1 failed link out of 2,159 evaluations"
}
],
"format": "epistemedia-quote-minimal-edition-projection-v0.1",
"license": "CC BY 4.0",
"license_treatment": "quote-minimal attributed spans",
"readback_receipts": [
{
"captured_bytes": 138385,
"captured_sha256": "7c5e3c33f3122b07d176e6679cd6762a95a6babaf4a353686f019f905b5526ca",
"edition_id": "edition:cnv-arxiv-v1",
"http_status": 200,
"media_type": "text/html; charset=utf-8",
"redirect_chain": [],
"requested_url": "https://arxiv.org/html/2605.06635",
"resolved_url": "https://arxiv.org/html/2605.06635",
"retrieval_status": "retrieved"
},
{
"captured_bytes": 138385,
"captured_sha256": "7c5e3c33f3122b07d176e6679cd6762a95a6babaf4a353686f019f905b5526ca",
"edition_id": "edition:cnv-arxiv-v1",
"http_status": 200,
"media_type": "text/html; charset=utf-8",
"redirect_chain": [],
"requested_url": "https://arxiv.org/html/2605.06635v1",
"resolved_url": "https://arxiv.org/html/2605.06635v1",
"retrieval_status": "retrieved"
},
{
"captured_bytes": 1424589,
"captured_sha256": "db5b7b7e3d9ce9fc6d6713b60f65f0d81b07090ded2a28be7f541d1c290230b5",
"edition_id": "edition:cnv-arxiv-v1",
"http_status": 200,
"media_type": "application/pdf",
"redirect_chain": [],
"requested_url": "https://arxiv.org/pdf/2605.06635",
"resolved_url": "https://arxiv.org/pdf/2605.06635",
"retrieval_status": "retrieved"
},
{
"captured_bytes": 43848,
"captured_sha256": "d1b6d476b3e81460dbc8821757775a8fa589e0579b77167ef0f33d35cb819cb5",
"edition_id": "edition:cnv-arxiv-v1",
"http_status": 200,
"media_type": "text/html; charset=utf-8",
"redirect_chain": [],
"requested_url": "https://arxiv.org/abs/2605.06635",
"resolved_url": "https://arxiv.org/abs/2605.06635",
"retrieval_status": "retrieved"
}
],
"review_supplements": [],
"source_text_file": "cited-not-verified.txt"
},
"content_digest": "sha256:90da1e5adccef8448150bd162daa5cdc08d01bfab5142d92de6aee947a79e58b",
"content_length": 7145,
"edition_label": "Quote-minimal projection of edition:cnv-arxiv-v1",
"id": "em:dossier-edition:sha256:601f0e4510819607265b94c8261884c072d559e63de9f71dbfc6bca5d33c5e29",
"key": "edition-cnv-arxiv-v1-f24001c50a",
"media_type": "application/json",
"retrieved_at": "2026-08-23T18:33:06Z",
"visibility": "public",
"work_key": "work-cited-not-verified-3825b25622"
},
{
"content": {
"canonical_url": "https://arxiv.org/html/2509.04499v1",
"excerpts": [
{
"license_treatment": "metadata, digest, and quote-minimal attributed spans only",
"locator": "Abstract, line 46",
"match_status": "exact-normalized-match",
"occurrence_ids": [
"V2-SOL-04:S4:S4-SP1"
],
"source_text_sha256": "d3d5e54884b46ff1815be6080405a9751e2b84b1a704a926e4af4806f68c9272",
"span_root_id": "span:sha256:8cce3d9ee1e1674ba1b93321dd9b99a55d834ca980451f8ab61bcaf5bdda8e4d",
"text": "citation accuracy ranging from 40–80% across systems"
},
{
"license_treatment": "metadata, digest, and quote-minimal attributed spans only",
"locator": "Section 3.2, lines 163-171",
"match_status": "exact-normalized-match",
"occurrence_ids": [
"V2-SOL-04:S4:S4-SP2"
],
"source_text_sha256": "d3d5e54884b46ff1815be6080405a9751e2b84b1a704a926e4af4806f68c9272",
"span_root_id": "span:sha256:9dd6eb46529e77a62748093528e52685fbd92d46a756aef750d22aa25981db23",
"text": "303 queries x 9 models"
}
],
"format": "epistemedia-quote-minimal-edition-projection-v0.1",
"license": "arXiv non-exclusive distribution license",
"license_treatment": "metadata, digest, and quote-minimal attributed spans only",
"readback_receipts": [
{
"captured_bytes": 364455,
"captured_sha256": "0c28edecaedd882584e985caac1e14f4a03dd59e939e17ae755a3c5e07ae42b0",
"edition_id": "edition:deeptrace-arxiv-v1",
"http_status": 200,
"media_type": "text/html; charset=utf-8",
"redirect_chain": [],
"requested_url": "https://arxiv.org/html/2509.04499",
"resolved_url": "https://arxiv.org/html/2509.04499",
"retrieval_status": "retrieved"
},
{
"captured_bytes": 364455,
"captured_sha256": "0c28edecaedd882584e985caac1e14f4a03dd59e939e17ae755a3c5e07ae42b0",
"edition_id": "edition:deeptrace-arxiv-v1",
"http_status": 200,
"media_type": "text/html; charset=utf-8",
"redirect_chain": [],
"requested_url": "https://arxiv.org/html/2509.04499v1",
"resolved_url": "https://arxiv.org/html/2509.04499v1",
"retrieval_status": "retrieved"
}
],
"review_supplements": [],
"source_text_file": "deeptrace-v1.txt"
},
"content_digest": "sha256:07a6f7e0282d888bc9157436b17fbde6286fb934b0efafd23399eebb7a22fa5d",
"content_length": 1935,
"edition_label": "Quote-minimal projection of edition:deeptrace-arxiv-v1",
"id": "em:dossier-edition:sha256:c7779c770aa24aa5925b5fe6e66f968c733e8522d30bcdc4290bc9409bbe8b5b",
"key": "edition-deeptrace-arxiv-v1-70a6e720c1",
"media_type": "application/json",
"retrieved_at": "2026-08-23T18:33:06Z",
"visibility": "public",
"work_key": "work-deeptrace-650d66cfcb"
},
{
"content": {
"canonical_url": "https://proceedings.iclr.cc/paper_files/paper/2026/file/ad08767706825033b99122332293033d-Paper-Conference.pdf",
"excerpts": [
{
"license_treatment": "metadata, digest, and quote-minimal attributed spans only",
"locator": "page 14, Table 3",
"match_status": "exact-normalized-match",
"occurrence_ids": [
"V2-SOL-01:s3_deeptrace:sp3d"
],
"source_text_sha256": "0eaa415eca25bed5e591b8250640d695b86f6dac6601f5f00167ee3ac7ab9171",
"span_root_id": "span:sha256:35927851edd2e7e5e0f60d98498940f88304ba99bf1f85a08535663943c2f40d",
"text": "Factual support (statement-source) 0.62 binary"
},
{
"license_treatment": "metadata, digest, and quote-minimal attributed spans only",
"locator": "pp.4-7, Sections 3.1.1 and 3.2",
"match_status": "exact-normalized-match",
"occurrence_ids": [
"V2-TERRA-02:s3_deeptrace:s3_corpus_and_retrieval"
],
"source_text_sha256": "0eaa415eca25bed5e591b8250640d695b86f6dac6601f5f00167ee3ac7ab9171",
"span_root_id": "span:sha256:9cb1700c08368feb234207045354cbf559d6ed80f2b0bae814d2b5b3c858b73c",
"text": "\"The dataset comprises 303 questions\""
},
{
"license_treatment": "metadata, digest, and quote-minimal attributed spans only",
"locator": "Official ICLR abstract",
"match_status": "exact-normalized-match",
"occurrence_ids": [
"V2-SOL-02:S6:S6_SPAN_1"
],
"source_text_sha256": "0eaa415eca25bed5e591b8250640d695b86f6dac6601f5f00167ee3ac7ab9171",
"span_root_id": "span:sha256:a08468cb62f2feb21d9cbb95175ce00f54592bd83573cc9166e2375947f62fee",
"text": "with citation accuracy ranging from 40–80% across systems"
},
{
"license_treatment": "metadata, digest, and quote-minimal attributed spans only",
"locator": "page 4, source scraping",
"match_status": "exact-normalized-match",
"occurrence_ids": [
"V2-SOL-01:s3_deeptrace:sp3c"
],
"source_text_sha256": "0eaa415eca25bed5e591b8250640d695b86f6dac6601f5f00167ee3ac7ab9171",
"span_root_id": "span:sha256:a15ab02e9fcb23df02fb03d28967d4220ddf65ad380d223049f22b6ab68ed5eb",
"text": "For roughly 15% of the URLs, the Reader tool returns an error"
},
{
"license_treatment": "metadata, digest, and quote-minimal attributed spans only",
"locator": "p.14, Table 3",
"match_status": "exact-normalized-match",
"occurrence_ids": [
"V2-TERRA-02:s3_deeptrace:s3_judge_validation"
],
"source_text_sha256": "0eaa415eca25bed5e591b8250640d695b86f6dac6601f5f00167ee3ac7ab9171",
"span_root_id": "span:sha256:aa1534ce2e177153ab85e512deb8d064d14ec1eae16d8032bdb5043252738d4c",
"text": "\"Factual support (statement–source) 0.62\""
},
{
"license_treatment": "metadata, digest, and quote-minimal attributed spans only",
"locator": "Paper section 3.1.1, factual-support validation",
"match_status": "exact-normalized-match",
"occurrence_ids": [
"V2-SOL-02:S6:S6_SPAN_2"
],
"source_text_sha256": "0eaa415eca25bed5e591b8250640d695b86f6dac6601f5f00167ee3ac7ab9171",
"span_root_id": "span:sha256:e2c9f6220843a04665f5d1cd142ac15f04211ace5c579fa67f77e74e84a76dfe",
"text": "Pearson correlation of 0.62 between the LLM judge and manual labels"
}
],
"format": "epistemedia-quote-minimal-edition-projection-v0.1",
"license": "unknown",
"license_treatment": "metadata, digest, and quote-minimal attributed spans only",
"readback_receipts": [
{
"captured_bytes": 10137,
"captured_sha256": "5457daae5b00c1c4105a9c80f5eb39a4027d79a70702071cc2892a00eb27e05d",
"edition_id": "edition:deeptrace-iclr-2026",
"http_status": 200,
"media_type": "text/html; charset=utf-8",
"redirect_chain": [],
"requested_url": "https://proceedings.iclr.cc/paper_files/paper/2026/hash/ad08767706825033b99122332293033d-Abstract-Conference.html",
"resolved_url": "https://proceedings.iclr.cc/paper_files/paper/2026/hash/ad08767706825033b99122332293033d-Abstract-Conference.html",
"retrieval_status": "retrieved"
},
{
"captured_bytes": 2982839,
"captured_sha256": "dea4981c1066d0240a005f603b3b14419c7e32beb5191b3df847ceb26af3d6b6",
"edition_id": "edition:deeptrace-iclr-2026",
"http_status": 200,
"media_type": "application/pdf",
"redirect_chain": [],
"requested_url": "https://proceedings.iclr.cc/paper_files/paper/2026/file/ad08767706825033b99122332293033d-Paper-Conference.pdf",
"resolved_url": "https://proceedings.iclr.cc/paper_files/paper/2026/file/ad08767706825033b99122332293033d-Paper-Conference.pdf",
"retrieval_status": "retrieved"
}
],
"review_supplements": [
{
"canonical_url": "https://proceedings.iclr.cc/paper_files/paper/2026/file/ad08767706825033b99122332293033d-Paper-Conference.pdf",
"carrier": {
"bytes": 2982839,
"media_type": "application/pdf",
"sha256": "dea4981c1066d0240a005f603b3b14419c7e32beb5191b3df847ceb26af3d6b6"
},
"edition_id": "edition:deeptrace-iclr-2026",
"extent": "Gemini(DR) demonstrates weak citation performance: only 40.3% citation accuracy",
"license_treatment": "metadata, digest, and quote-minimal attributed spans only",
"locator": "printed page 10; PDF page index 10/22; Section 4 Results; Deep Research Agents paragraph",
"role": "qualification-only",
"supplement_id": "supplement:deeptrace-gemini-prose-40-3",
"warrant_id": "warrant:deeptrace-support-variation",
"work_id": "work:deeptrace"
},
{
"canonical_url": "https://proceedings.iclr.cc/paper_files/paper/2026/file/ad08767706825033b99122332293033d-Paper-Conference.pdf",
"carrier": {
"bytes": 2982839,
"media_type": "application/pdf",
"sha256": "dea4981c1066d0240a005f603b3b14419c7e32beb5191b3df847ceb26af3d6b6"
},
"edition_id": "edition:deeptrace-iclr-2026",
"extent": {
"cell": "50.3",
"column": "Gemini (DR)",
"row": "%Citation Accuracy"
},
"license_treatment": "metadata, digest, and quote-minimal attributed spans only",
"locator": "printed page 9; PDF page index 9/22; Table 1; Gemini (DR) column; %Citation Accuracy row",
"role": "qualification-only",
"supplement_id": "supplement:deeptrace-gemini-table-50-3",
"warrant_id": "warrant:deeptrace-support-variation",
"work_id": "work:deeptrace"
}
],
"source_text_file": "deeptrace-iclr.txt"
},
"content_digest": "sha256:0e0c4dd57c80c2ab32707ec1b17eda3ea21fc75ec4ce5191875446912649b545",
"content_length": 5645,
"edition_label": "Quote-minimal projection of edition:deeptrace-iclr-2026",
"id": "em:dossier-edition:sha256:1a587168a59e161ea083b7b289339c5f41444a26f2f82a2eb8acb43b27367f70",
"key": "edition-deeptrace-iclr-2026-b2b153eaa6",
"media_type": "application/json",
"retrieved_at": "2026-08-23T18:33:06Z",
"visibility": "public",
"work_key": "work-deeptrace-650d66cfcb"
},
{
"content": {
"canonical_url": "https://arxiv.org/html/2506.11763v1",
"excerpts": [
{
"license_treatment": "quote-minimal attributed spans",
"locator": "Evaluation methodology, lines 144-149",
"match_status": "exact-normalized-match",
"occurrence_ids": [
"V2-TERRA-03:S2:S2-A"
],
"source_text_sha256": "c8391dc3b757d97aa6d781636819180d34cb7343d290896d54db43f6f467da89",
"span_root_id": "span:sha256:015bb68c904bf225e84c82b1acffb3cc38a4a087d7b0f79d3b0102b685b2265e",
"text": "Each unique Statement-URL pair undergoes a support evaluation."
},
{
"license_treatment": "quote-minimal attributed spans",
"locator": "Table 1, lines 177-182",
"match_status": "ordered-fragment-match",
"occurrence_ids": [
"V2-TERRA-03:S2:S2-B"
],
"source_text_sha256": "c8391dc3b757d97aa6d781636819180d34cb7343d290896d54db43f6f467da89",
"span_root_id": "span:sha256:0d4f357f2a5565bdaea01a1e537650958d71216868d1a2b0bc14c47a45b6d0ef",
"text": "Perplexity Deep Research | ... | 90.24 | 31.26 ... Gemini-2.5-Pro Deep Research | ... | 81.44 | 111.21 ... OpenAI Deep Research | ... | 77.96 | 40.79"
},
{
"license_treatment": "quote-minimal attributed spans",
"locator": "Section 3.2, lines 144-146",
"match_status": "exact-normalized-match",
"occurrence_ids": [
"V2-SOL-04:S2:S2-SP2"
],
"source_text_sha256": "c8391dc3b757d97aa6d781636819180d34cb7343d290896d54db43f6f467da89",
"span_root_id": "span:sha256:13cb0ddb087b82988b7445a9cfb02562f9c044ef4f244e8b176ff475859eaea3",
"text": "‘support’ or ‘not support’"
},
{
"license_treatment": "quote-minimal attributed spans",
"locator": "Table 1, Perplexity Deep Research citation-accuracy cell",
"match_status": "exact-normalized-match",
"occurrence_ids": [
"V2-SOL-04:S2:S2-SP1"
],
"source_text_sha256": "c8391dc3b757d97aa6d781636819180d34cb7343d290896d54db43f6f467da89",
"span_root_id": "span:sha256:196c54d6212fbc0c054dbf8e8467f2388c1f788bb93f4b1a4b6ae085eed17685",
"text": "90.24"
},
{
"license_treatment": "quote-minimal attributed spans",
"locator": "Table 1, OpenAI Deep Research row; FACT columns",
"match_status": "exact-normalized-match",
"occurrence_ids": [
"V2-SOL-03:s2:s2_span1"
],
"source_text_sha256": "c8391dc3b757d97aa6d781636819180d34cb7343d290896d54db43f6f467da89",
"span_root_id": "span:sha256:2f311bfb451c6208101cb7583af37c338914fcdb10f5c75b131e23d5bb08e66a",
"text": "OpenAI Deep Research | 46.98 | 46.87 | 45.25 | 49.27 | 47.14 | 77.96 | 40.79"
},
{
"license_treatment": "quote-minimal attributed spans",
"locator": "Appendix D, Table 5",
"match_status": "exact-normalized-match",
"occurrence_ids": [
"V2-TERRA-04:S1_deepresearchbench:S1_time"
],
"source_text_sha256": "c8391dc3b757d97aa6d781636819180d34cb7343d290896d54db43f6f467da89",
"span_root_id": "span:sha256:76e9bd59b31c7ebd2ed7a2186e59753a79d3c730e0d7b4fbf0146fb39f4e9c5f",
"text": "\"OpenAI Deep Research | April 1 – May 8\""
},
{
"license_treatment": "quote-minimal attributed spans",
"locator": "Section 3.2, paragraphs 'Statement-URL Pair Extraction and Deduplication' and 'Support Judgment'",
"match_status": "exact-normalized-match",
"occurrence_ids": [
"V2-TERRA-04:S1_deepresearchbench:S1_method"
],
"source_text_sha256": "c8391dc3b757d97aa6d781636819180d34cb7343d290896d54db43f6f467da89",
"span_root_id": "span:sha256:9fe42cb48702d8d96ebbbc0309692214ece6ec495c447c06e2be62f32ff00894",
"text": "\"Each unique Statement-URL pair undergoes a support evaluation.\""
},
{
"license_treatment": "quote-minimal attributed spans",
"locator": "p. 4, lines 150-159",
"match_status": "ordered-fragment-match",
"occurrence_ids": [
"V2-TERRA-01:s1_deepresearchbench:s1_method"
],
"source_text_sha256": "c8391dc3b757d97aa6d781636819180d34cb7343d290896d54db43f6f467da89",
"span_root_id": "span:sha256:b0a39c963cdfdc5f9852cfc9f7e33e05d31232bd1d8998851e879c38e6c34843",
"text": "\"Each unique Statement-URL pair undergoes a support evaluation... This results in a binary judgment ('support' or 'not support') for each pair, determining whether the citation accurately grounds the claim.\""
},
{
"license_treatment": "quote-minimal attributed spans",
"locator": "Appendix C, line 365",
"match_status": "exact-normalized-match",
"occurrence_ids": [
"V2-SOL-04:S2:S2-SP3"
],
"source_text_sha256": "c8391dc3b757d97aa6d781636819180d34cb7343d290896d54db43f6f467da89",
"span_root_id": "span:sha256:b83996e3b5fdf878e04d6d41d0e7a1eebee9fdd9bbccf2ae4eaf33524fcfca44",
"text": "96% of cases"
},
{
"license_treatment": "quote-minimal attributed spans",
"locator": "pp.17-18, Appendix E, equations 4-6",
"match_status": "exact-normalized-match",
"occurrence_ids": [
"V2-TERRA-02:s1_deepresearchbench:s1_fact_formula"
],
"source_text_sha256": "c8391dc3b757d97aa6d781636819180d34cb7343d290896d54db43f6f467da89",
"span_root_id": "span:sha256:e4a9f30effc031388ed5c9993dc6f071903bec5e3f1998db4c08ab00b284b27f",
"text": "\"the proportion of 'support' statement-URL pairs for each individual task\""
},
{
"license_treatment": "quote-minimal attributed spans",
"locator": "p.4, Section 3.2, lines 148-155",
"match_status": "exact-normalized-match",
"occurrence_ids": [
"V2-TERRA-02:s1_deepresearchbench:s1_fact_method"
],
"source_text_sha256": "c8391dc3b757d97aa6d781636819180d34cb7343d290896d54db43f6f467da89",
"span_root_id": "span:sha256:eaf37bef7b8dc35f3e00b85d3e58ad872dc0ea604a062c8cdd8929184587d5d9",
"text": "\"Each unique Statement-URL pair undergoes a support evaluation.\""
},
{
"license_treatment": "quote-minimal attributed spans",
"locator": "Appendix C, line 365",
"match_status": "exact-normalized-match",
"occurrence_ids": [
"V2-SOL-04:S2:S2-SP4"
],
"source_text_sha256": "c8391dc3b757d97aa6d781636819180d34cb7343d290896d54db43f6f467da89",
"span_root_id": "span:sha256:f35c9565f2fd047f6e7272dc608bbf6a63813a717bad5354c24bdaac65e40bfe",
"text": "92% of cases"
},
{
"license_treatment": "quote-minimal attributed spans",
"locator": "p.17, Appendix C, lines 704-708",
"match_status": "exact-normalized-match",
"occurrence_ids": [
"V2-TERRA-02:s1_deepresearchbench:s1_fact_judge_validation"
],
"source_text_sha256": "c8391dc3b757d97aa6d781636819180d34cb7343d290896d54db43f6f467da89",
"span_root_id": "span:sha256:f98f1274859ba98cc538e3be6d11d345a5382cfcb9516aa1cf3f3278ea646b92",
"text": "\"aligned with human 'support' determinations in 96% of cases\""
}
],
"format": "epistemedia-quote-minimal-edition-projection-v0.1",
"license": "CC BY 4.0",
"license_treatment": "quote-minimal attributed spans",
"readback_receipts": [
{
"captured_bytes": 352002,
"captured_sha256": "9aa2894dbeaac30b23e7ffc8107a7f53b6e3855c8511838551de5bd7a422cc42",
"edition_id": "edition:drbench-arxiv-v1",
"http_status": 200,
"media_type": "text/html; charset=utf-8",
"redirect_chain": [],
"requested_url": "https://arxiv.org/html/2506.11763",
"resolved_url": "https://arxiv.org/html/2506.11763",
"retrieval_status": "retrieved"
},
{
"captured_bytes": 352002,
"captured_sha256": "9aa2894dbeaac30b23e7ffc8107a7f53b6e3855c8511838551de5bd7a422cc42",
"edition_id": "edition:drbench-arxiv-v1",
"http_status": 200,
"media_type": "text/html; charset=utf-8",
"redirect_chain": [],
"requested_url": "https://arxiv.org/html/2506.11763v1",
"resolved_url": "https://arxiv.org/html/2506.11763v1",
"retrieval_status": "retrieved"
},
{
"captured_bytes": 3678399,
"captured_sha256": "8fbf30398f5e62f8839f0c9c8609bbb9e3cd0b57ae27d4bf33cb5db2007d1118",
"edition_id": "edition:drbench-arxiv-v1",
"http_status": 200,
"media_type": "application/pdf",
"redirect_chain": [],
"requested_url": "https://deepresearch-bench.github.io/static/papers/deepresearch-bench.pdf",
"resolved_url": "https://deepresearch-bench.github.io/static/papers/deepresearch-bench.pdf",
"retrieval_status": "retrieved"
}
],
"review_supplements": [],
"source_text_file": "drbench-v1.txt"
},
"content_digest": "sha256:023aac7d22e7ca1230e2ffa237a4ea24c9f69f8b27653ddb6a494d4a30e4331c",
"content_length": 7381,
"edition_label": "Quote-minimal projection of edition:drbench-arxiv-v1",
"id": "em:dossier-edition:sha256:c0f1d83abed0c9adcc86259b5bfe38e3f18b290f25227b5b9e77c99322fe0b2c",
"key": "edition-drbench-arxiv-v1-7afa5ade36",
"media_type": "application/json",
"retrieved_at": "2026-08-23T18:33:06Z",
"visibility": "public",
"work_key": "work-deepresearch-bench-paper-8ef36010d2"
},
{
"content": {
"canonical_url": "https://proceedings.iclr.cc/paper_files/paper/2026/file/465f22be10e07b301c6ed58f0472f704-Paper-Conference.pdf",
"excerpts": [
{
"license_treatment": "metadata, digest, and quote-minimal attributed spans only",
"locator": "page 16, FACT judge validation",
"match_status": "exact-normalized-match",
"occurrence_ids": [
"V2-SOL-01:s2_drbench:sp2d"
],
"source_text_sha256": "ced9e7b0ae4ab2c26add3230a7b40200673ee4fc0c5c763df226bfe784d60016",
"span_root_id": "span:sha256:279db725380c133044944dccb379b75a396242219aa30437b26ddc40b09a3b5c",
"text": "aligned with human 'support' determinations in 96% of cases"
},
{
"license_treatment": "metadata, digest, and quote-minimal attributed spans only",
"locator": "page 4, Section 3.2, Support Judgment",
"match_status": "exact-normalized-match",
"occurrence_ids": [
"V2-SOL-01:s2_drbench:sp2a"
],
"source_text_sha256": "ced9e7b0ae4ab2c26add3230a7b40200673ee4fc0c5c763df226bfe784d60016",
"span_root_id": "span:sha256:7d7331942f9d0520e18e96206e7718e5187788be4d48cef995fac79b235db757",
"text": "This yields a binary judgment: 'support' or 'not support'."
},
{
"license_treatment": "metadata, digest, and quote-minimal attributed spans only",
"locator": "page 5, Section 4.1",
"match_status": "exact-normalized-match",
"occurrence_ids": [
"V2-SOL-01:s2_drbench:sp2c"
],
"source_text_sha256": "ced9e7b0ae4ab2c26add3230a7b40200673ee4fc0c5c763df226bfe784d60016",
"span_root_id": "span:sha256:c27aaca02bda06ed764be53351158fc862af9d4a556d3e82074507436f48c3f8",
"text": "complete set of 100 tasks"
}
],
"format": "epistemedia-quote-minimal-edition-projection-v0.1",
"license": "unknown",
"license_treatment": "metadata, digest, and quote-minimal attributed spans only",
"readback_receipts": [
{
"captured_bytes": 12692,
"captured_sha256": "a2d811e0d24af89d9bd4b19cc9e8fbc474c67ef3827648550a6560eca2a5e332",
"edition_id": "edition:drbench-iclr-2026",
"http_status": 403,
"media_type": "text/html",
"redirect_chain": [],
"requested_url": "https://openreview.net/pdf?id=hQ0K2Hhq7H",
"resolved_url": "https://openreview.net/pdf?id=hQ0K2Hhq7H",
"retrieval_status": "inaccessible"
},
{
"captured_bytes": 2665857,
"captured_sha256": "8f80ce247f7cc355bb6773f36037e01cc3ba2e2082c77381eaadf0d2b92b021c",
"edition_id": "edition:drbench-iclr-2026",
"http_status": 200,
"media_type": "application/pdf",
"redirect_chain": [],
"requested_url": "https://proceedings.iclr.cc/paper_files/paper/2026/file/465f22be10e07b301c6ed58f0472f704-Paper-Conference.pdf",
"resolved_url": "https://proceedings.iclr.cc/paper_files/paper/2026/file/465f22be10e07b301c6ed58f0472f704-Paper-Conference.pdf",
"retrieval_status": "retrieved"
}
],
"review_supplements": [],
"source_text_file": "drbench-iclr.txt"
},
"content_digest": "sha256:02f53bf5ae17ab2bffc25ba8dd5ca912d35b85ae6ae78a202cd79fe7f7efe25f",
"content_length": 2610,
"edition_label": "Quote-minimal projection of edition:drbench-iclr-2026",
"id": "em:dossier-edition:sha256:47d8d33326a498097021dc5754365cbb8f5abfd3e628e90908d2342f0b6c7cf1",
"key": "edition-drbench-iclr-2026-d00a3dcdb6",
"media_type": "application/json",
"retrieved_at": "2026-08-23T18:33:06Z",
"visibility": "public",
"work_key": "work-deepresearch-bench-paper-8ef36010d2"
},
{
"content": {
"canonical_url": "https://github.com/Ayanami0730/deep_research_bench/commit/469cce54ea7f6a63c163d3d9fec879cf289ec484",
"excerpts": [
{
"license_treatment": "metadata and quote-minimal README spans",
"locator": "repository README, Overview",
"match_status": "exact-normalized-match",
"occurrence_ids": [
"V2-SOL-01:s2a_drbench_repo:sp2e"
],
"source_text_sha256": "afebe200cb45042b2783d6ad581770bbd8674027d4e7443e2eb6c26720c6cb0c",
"span_root_id": "span:sha256:0abbab3d6269d58eb76d8e90eff1e4c27512cf46d9b8e5ff0a9f537b18f232fa",
"text": "100 PhD-level research tasks"
},
{
"license_treatment": "metadata and quote-minimal README spans",
"locator": "README FACT, lines 240-248",
"match_status": "exact-normalized-match",
"occurrence_ids": [
"V2-TERRA-03:S5:S5-B"
],
"source_text_sha256": "afebe200cb45042b2783d6ad581770bbd8674027d4e7443e2eb6c26720c6cb0c",
"span_root_id": "span:sha256:92eb7cf19745e24190dda84c77d590941134d09f06805c0fb605207fd942ae11",
"text": "Support Verification: Uses web scraping and LLM judgment to verify whether cited sources actually support the claims"
},
{
"license_treatment": "metadata and quote-minimal README spans",
"locator": "README News, 2026-05-11 evaluator migration notice",
"match_status": "exact-normalized-match",
"occurrence_ids": [
"V2-SOL-03:s7:s7_span1"
],
"source_text_sha256": "afebe200cb45042b2783d6ad581770bbd8674027d4e7443e2eb6c26720c6cb0c",
"span_root_id": "span:sha256:e5bcf0241a1aeae7d9a92755b1bcf0adeeea93a8922f22f832693c61780386e7",
"text": "Legacy code: the previous Gemini-2.5-Pro / Gemini-2.5-Flash evaluation code is preserved on the Gemini-2.5 branch."
},
{
"license_treatment": "metadata and quote-minimal README spans",
"locator": "README News, lines 176-188",
"match_status": "ordered-fragment-match",
"occurrence_ids": [
"V2-TERRA-03:S5:S5-A"
],
"source_text_sha256": "afebe200cb45042b2783d6ad581770bbd8674027d4e7443e2eb6c26720c6cb0c",
"span_root_id": "span:sha256:ef637b38f36195bea09205d532e8744daf3e257f93d97144f061024092bf67fe",
"text": "Official Evaluator Switched to GPT-5.5 ... Legacy code: the previous Gemini-2.5-Pro / Gemini-2.5-Flash evaluation code is preserved"
}
],
"format": "epistemedia-quote-minimal-edition-projection-v0.1",
"license": "Apache-2.0",
"license_treatment": "metadata and quote-minimal README spans",
"readback_receipts": [
{
"captured_bytes": 373533,
"captured_sha256": "ec9a4efdb736c6b5ec198cfbc5534eb67b5d421efa12cf795d7990be6c1e3b54",
"edition_id": "edition:drbench-repo-main-469cce5",
"http_status": 200,
"media_type": "text/html; charset=utf-8",
"redirect_chain": [],
"requested_url": "https://github.com/Ayanami0730/deep_research_bench",
"resolved_url": "https://github.com/Ayanami0730/deep_research_bench",
"retrieval_status": "retrieved"
}
],
"review_supplements": [],
"source_text_file": "drbench-repo.txt"
},
"content_digest": "sha256:35350075b3514f668c0deddb5bd4850ec31c7bfda2cf1e363995d68c580fe81b",
"content_length": 2699,
"edition_label": "Quote-minimal projection of edition:drbench-repo-main-469cce5",
"id": "em:dossier-edition:sha256:bf5c65b640b321fb75bf8b7f96854f151f834a7d31a2fc9d2bfb783852e7262b",
"key": "edition-drbench-repo-main-469cce5-4f8964f3e3",
"media_type": "application/json",
"retrieved_at": "2026-08-23T18:33:06Z",
"visibility": "public",
"work_key": "work-deepresearch-bench-repository-1e475c7631"
},
{
"content": {
"canonical_url": "https://data.mendeley.com/datasets/3s73z9zf3c/1",
"excerpts": [
{
"license_treatment": "metadata and quote-minimal landing-page spans",
"locator": "Dataset description",
"match_status": "exact-normalized-match",
"occurrence_ids": [
"V2-SOL-04:S6:S6-SP1"
],
"source_text_sha256": "1ed2ccfeb91e0346787a5c3122e191a6798d522a86f740e10ae9956c5b3f4cd4",
"span_root_id": "span:sha256:94ad583514aa7e04b8d5f7b8f9cdda6fdffe8ca6712f033da3ce4f18fe7c03f1",
"text": "sentence-level evaluation"
}
],
"format": "epistemedia-quote-minimal-edition-projection-v0.1",
"license": "CC BY 4.0",
"license_treatment": "metadata and quote-minimal landing-page spans",
"readback_receipts": [
{
"captured_bytes": 115612,
"captured_sha256": "da6308937337b2dfd3f97dc62f817e4a7ec362329ede0fce4c9bcdfd99ad93b1",
"edition_id": "edition:keplinger-supplement-v1",
"http_status": 200,
"media_type": "text/html; charset=utf-8",
"redirect_chain": [],
"requested_url": "https://data.mendeley.com/datasets/3s73z9zf3c/1",
"resolved_url": "https://data.mendeley.com/datasets/3s73z9zf3c/1",
"retrieval_status": "retrieved"
}
],
"review_supplements": [],
"source_text_file": "mendeley-v1.txt"
},
"content_digest": "sha256:0a485e13dcc56e5f883213fde7e24bdcc4ffb810bda8ac95186f83f6325fafd1",
"content_length": 1112,
"edition_label": "Quote-minimal projection of edition:keplinger-supplement-v1",
"id": "em:dossier-edition:sha256:8c80a50cd40cbb31bdaedb6730e8e173e3edc0f3440538dbcc4878eb0997ffd9",
"key": "edition-keplinger-supplement-v1-f6ddb4c04a",
"media_type": "application/json",
"retrieved_at": "2026-08-23T18:33:06Z",
"visibility": "public",
"work_key": "work-keplinger-supplement-1a03034fb3"
},
{
"content": {
"canonical_url": "https://data.mendeley.com/datasets/3s73z9zf3c/2",
"excerpts": [
{
"license_treatment": "metadata and quote-minimal landing-page spans; file API required authentication and was not used",
"locator": "dataset page, Description",
"match_status": "exact-normalized-match",
"occurrence_ids": [
"V2-SOL-01:s1a_keplinger_data:sp1d"
],
"source_text_sha256": "2e96cc4ae8e28c60fcb94dd40ef64a15252e4416e7bf33185b8b76ea6ff228e4",
"span_root_id": "span:sha256:6f79c3c824ccf889d73e234335661e5fca68d90f8c78ee97aef488e4fdb32a49",
"text": "Prompts for Deep Research-generated reviews and evaluation results"
}
],
"format": "epistemedia-quote-minimal-edition-projection-v0.1",
"license": "CC BY 4.0",
"license_treatment": "metadata and quote-minimal landing-page spans; file API required authentication and was not used",
"readback_receipts": [
{
"captured_bytes": 116951,
"captured_sha256": "0a2c4c6ed54dd7e2e0ca6cc07aa19b18fa1439e506c2310db778094282d2cebc",
"edition_id": "edition:keplinger-supplement-v2",
"http_status": 200,
"media_type": "text/html; charset=utf-8",
"redirect_chain": [],
"requested_url": "https://data.mendeley.com/datasets/3s73z9zf3c/2",
"resolved_url": "https://data.mendeley.com/datasets/3s73z9zf3c/2",
"retrieval_status": "retrieved"
}
],
"review_supplements": [],
"source_text_file": "mendeley-v2-page.txt"
},
"content_digest": "sha256:945e14af08757c25cdbcfe2d9724c092f2cbc668a2a9ff0d991dbedf4d12277b",
"content_length": 1280,
"edition_label": "Quote-minimal projection of edition:keplinger-supplement-v2",
"id": "em:dossier-edition:sha256:dca82f9514bbd76d7afb655b2fe3d11d98dc6950b90c3065d81afa335bd86418",
"key": "edition-keplinger-supplement-v2-01e44c09d8",
"media_type": "application/json",
"retrieved_at": "2026-08-23T18:33:06Z",
"visibility": "public",
"work_key": "work-keplinger-supplement-1a03034fb3"
},
{
"content": {
"canonical_url": "https://pmc.ncbi.nlm.nih.gov/articles/PMC13109748/",
"excerpts": [
{
"license_treatment": "quote-minimal attributed spans; no full-text redistribution",
"locator": "Main text, paragraph beginning 'For their high performances in generating reference lists'",
"match_status": "exact-normalized-match",
"occurrence_ids": [
"V2-SOL-02:S2:S2_SPAN_1"
],
"source_text_sha256": "8fd80ac582690bc0733dec0ebd65e9894a05915d117384f3e8bb6d8b0840a7c1",
"span_root_id": "span:sha256:092745c7db6407d4d52927642741cde229fad05a5347bb1867a5154421060e70",
"text": "citation-bearing sentences exhibited high error rates: 51.3 ± 6.5% for ChatGPT and 57.8 ± 22.7% for Le Chat"
},
{
"license_treatment": "quote-minimal attributed spans; no full-text redistribution",
"locator": "Table 1, ChatGPT all-correct subtotal",
"match_status": "exact-normalized-match",
"occurrence_ids": [
"V2-SOL-04:S5:S5-SP2"
],
"source_text_sha256": "8fd80ac582690bc0733dec0ebd65e9894a05915d117384f3e8bb6d8b0840a7c1",
"span_root_id": "span:sha256:39a9931f05ea356fe904ed15768fa08c66440b1fcb20f98e267a315e6d562af9",
"text": "16 (69.6%)"
},
{
"license_treatment": "quote-minimal attributed spans; no full-text redistribution",
"locator": "main text, paragraph beginning 'For their high performances'; Figure 1",
"match_status": "exact-normalized-match",
"occurrence_ids": [
"V2-SOL-03:s1:s1_span1"
],
"source_text_sha256": "8fd80ac582690bc0733dec0ebd65e9894a05915d117384f3e8bb6d8b0840a7c1",
"span_root_id": "span:sha256:46c392bc942cd88d525d74298e6bc6ebd9aeaec533ddfb7b7b761a9f5b454b05",
"text": "citation-bearing sentences exhibited high error rates: 51.3 ± 6.5% for ChatGPT and 57.8 ± 22.7% for Le Chat"
},
{
"license_treatment": "quote-minimal attributed spans; no full-text redistribution",
"locator": "Table 1, ChatGPT subtotal",
"match_status": "exact-normalized-match",
"occurrence_ids": [
"V2-SOL-04:S5:S5-SP1"
],
"source_text_sha256": "8fd80ac582690bc0733dec0ebd65e9894a05915d117384f3e8bb6d8b0840a7c1",
"span_root_id": "span:sha256:8c8d9a6dc003bf0298542b85df3f6f08db631e68e89f606f4e53e36e86dde167",
"text": "23 (100.0%)"
},
{
"license_treatment": "quote-minimal attributed spans; no full-text redistribution",
"locator": "Main text, paragraph beginning 'To address the gap' (search-indexed primary full text)",
"match_status": "exact-normalized-match",
"occurrence_ids": [
"V2-TERRA-03:S4:S4-A"
],
"source_text_sha256": "8fd80ac582690bc0733dec0ebd65e9894a05915d117384f3e8bb6d8b0840a7c1",
"span_root_id": "span:sha256:b38a2af5982b4c85bc205d2f533a23ed3a8f40a49b651a5f00ec6932dcc67f7d",
"text": "We imposed a 1500-word limit, APA-style references and conducted three independent runs."
},
{
"license_treatment": "quote-minimal attributed spans; no full-text redistribution",
"locator": "Main text, following paragraph (search-indexed primary full text)",
"match_status": "exact-normalized-match",
"occurrence_ids": [
"V2-TERRA-03:S4:S4-B"
],
"source_text_sha256": "8fd80ac582690bc0733dec0ebd65e9894a05915d117384f3e8bb6d8b0840a7c1",
"span_root_id": "span:sha256:b6a6ba48a2c4e352c8565b671a15a90400739f2aeefda8a65a31ad8a1ff9a4e4",
"text": "ChatGPT Deep Research-generated references were mostly identifiable by their metadata (95.7%): 69.6% were entirely correct"
},
{
"license_treatment": "quote-minimal attributed spans; no full-text redistribution",
"locator": "body, paragraph immediately after Table 1",
"match_status": "exact-normalized-match",
"occurrence_ids": [
"V2-SOL-01:s1_keplinger:sp1b"
],
"source_text_sha256": "8fd80ac582690bc0733dec0ebd65e9894a05915d117384f3e8bb6d8b0840a7c1",
"span_root_id": "span:sha256:b903dc2284e1b4d80bb6ef948d79bbdbc7d7447e871281b8577284883f75ccfa",
"text": "high error rates: 51.3 ± 6.5% for ChatGPT and 57.8 ± 22.7% for Le Chat"
},
{
"license_treatment": "quote-minimal attributed spans; no full-text redistribution",
"locator": "Paragraph after Table 1",
"match_status": "exact-normalized-match",
"occurrence_ids": [
"V2-SOL-04:S5:S5-SP3"
],
"source_text_sha256": "8fd80ac582690bc0733dec0ebd65e9894a05915d117384f3e8bb6d8b0840a7c1",
"span_root_id": "span:sha256:f05b3189dc3fafd45b5bde643a91fd0616832928215a37a2430a5a658e592029",
"text": "51.3 ± 6.5%"
}
],
"format": "epistemedia-quote-minimal-edition-projection-v0.1",
"license": "CC BY-NC 4.0",
"license_treatment": "quote-minimal attributed spans; no full-text redistribution",
"readback_receipts": [
{
"captured_bytes": 111993,
"captured_sha256": "abbaa9f22035230d13152a68aaeecd8e83e40867c9dae4acd1d547e36c866793",
"edition_id": "edition:keplinger-vor-2025",
"http_status": 200,
"media_type": "text/html; charset=utf-8",
"redirect_chain": [],
"requested_url": "https://pmc.ncbi.nlm.nih.gov/articles/PMC13109748/",
"resolved_url": "https://pmc.ncbi.nlm.nih.gov/articles/PMC13109748/",
"retrieval_status": "retrieved"
},
{
"captured_bytes": 5498,
"captured_sha256": "d74c1891ae09a2b89121f2a60ac837c89761a8086759e45db214ca68c1905cea",
"edition_id": "edition:keplinger-vor-2025",
"http_status": 403,
"media_type": "text/html; charset=UTF-8",
"redirect_chain": [],
"requested_url": "https://onlinelibrary.wiley.com/doi/10.1111/jdv.70035",
"resolved_url": "https://onlinelibrary.wiley.com/doi/10.1111/jdv.70035",
"retrieval_status": "inaccessible"
},
{
"captured_bytes": 5565,
"captured_sha256": "a46109544fe4ff4504fab5e97abea3cb7172367aba54a308937386394f0ff046",
"edition_id": "edition:keplinger-vor-2025",
"http_status": 203,
"media_type": "text/html; charset=utf-8",
"redirect_chain": [],
"requested_url": "https://pubmed.ncbi.nlm.nih.gov/40904191/",
"resolved_url": "https://pubmed.ncbi.nlm.nih.gov/40904191/",
"retrieval_status": "inaccessible"
}
],
"review_supplements": [],
"source_text_file": "keplinger.txt"
},
"content_digest": "sha256:98431c96e776a0502a6387574529817629ca50e164d8cd4e7f6997215ec7e097",
"content_length": 5462,
"edition_label": "Quote-minimal projection of edition:keplinger-vor-2025",
"id": "em:dossier-edition:sha256:ecbbc5366e2c53ebf6358f0f4c06b7076e158210edd3a081f16af7ab0b029761",
"key": "edition-keplinger-vor-2025-10ba10f860",
"media_type": "application/json",
"retrieved_at": "2026-08-23T18:33:06Z",
"visibility": "public",
"work_key": "work-keplinger-dermatology-audit-639f9ea37a"
},
{
"content": {
"canonical_url": "https://arxiv.org/pdf/2510.14240v5",
"excerpts": [
{
"license_treatment": "quote-minimal attributed spans; raw CC BY claim is corrected in review records",
"locator": "p.33, Appendix E",
"match_status": "ordered-fragment-match",
"occurrence_ids": [
"V2-TERRA-02:s2_liveresearchbench:s2_rubric_tree"
],
"source_text_sha256": "6605fcf67de305d0313f395e2493bf6c3bbd7ae87d7fa08358364ded225d9b77",
"span_root_id": "span:sha256:d880991ad784390029d49f84949d083c4789d67d45893d87cfb6b5a38703453f",
"text": "\"E1: URL is inaccessible or does not resolve; E2: URL content is irrelevant ... E3: URL content does not support the specific statements.\""
},
{
"license_treatment": "quote-minimal attributed spans; raw CC BY claim is corrected in review records",
"locator": "p.23, Appendix C, Citation Accuracy",
"match_status": "exact-normalized-match",
"occurrence_ids": [
"V2-TERRA-02:s2_liveresearchbench:s2_validation"
],
"source_text_sha256": "6605fcf67de305d0313f395e2493bf6c3bbd7ae87d7fa08358364ded225d9b77",
"span_root_id": "span:sha256:e6f447129d9a662fb202d4ac22e16822bacfdbcbcdb7afe3b589e21802890b8a",
"text": "\"human evaluators agree with either Gemini or GPT-5 on 87.1%\""
}
],
"format": "epistemedia-quote-minimal-edition-projection-v0.1",
"license": "CC BY-NC-SA 4.0",
"license_treatment": "quote-minimal attributed spans; raw CC BY claim is corrected in review records",
"readback_receipts": [
{
"captured_bytes": 13057891,
"captured_sha256": "579b9728b76cfd242e9c94d9ff2985e196bbc72b5a741030e4f308ede04a4f69",
"edition_id": "edition:liveresearchbench-arxiv-v5",
"http_status": 200,
"media_type": "application/pdf",
"redirect_chain": [],
"requested_url": "https://arxiv.org/pdf/2510.14240v5",
"resolved_url": "https://arxiv.org/pdf/2510.14240v5",
"retrieval_status": "retrieved"
}
],
"review_supplements": [],
"source_text_file": "liveresearchbench.txt"
},
"content_digest": "sha256:57db6f35bb09eae11b6993da86a1eb3a898d604a15466d02e40d2e7efeb803dd",
"content_length": 1801,
"edition_label": "Quote-minimal projection of edition:liveresearchbench-arxiv-v5",
"id": "em:dossier-edition:sha256:73c767cab5707e7c87ea4cb96cc6a3434512ae75363371f6d0aad2a365c4e11c",
"key": "edition-liveresearchbench-arxiv-v5-f5644763e5",
"media_type": "application/json",
"retrieved_at": "2026-08-23T18:33:06Z",
"visibility": "public",
"work_key": "work-liveresearchbench-61e7da557b"
},
{
"content": {
"canonical_url": "https://arxiv.org/html/2508.15804v1",
"excerpts": [
{
"license_treatment": "quote-minimal attributed spans",
"locator": "Appendix A.1 'Limitations'",
"match_status": "exact-normalized-match",
"occurrence_ids": [
"V2-TERRA-04:S3_reportbench:S3_limitations"
],
"source_text_sha256": "014f247563e015dfe7cd6ea4dee6c410f57c74788a6fb5bd589f65df078e8c95",
"span_root_id": "span:sha256:084f11e57cf779850ea87aaf30ba1277438ff9efcb533fd2ebb09dba3665be8d",
"text": "\"most of which are concentrated in STEM fields\""
},
{
"license_treatment": "quote-minimal attributed spans",
"locator": "Limitations, lines 259-261",
"match_status": "exact-normalized-match",
"occurrence_ids": [
"V2-TERRA-03:S3:S3-D"
],
"source_text_sha256": "014f247563e015dfe7cd6ea4dee6c410f57c74788a6fb5bd589f65df078e8c95",
"span_root_id": "span:sha256:1db12b68cfeb4c5f5a96fa9cd2e2bd46e8dfb4f096f73154481d29e22a4d5414",
"text": "The benchmark primarily draws from peer-reviewed survey papers on arXiv, most of which are concentrated in STEM fields."
},
{
"license_treatment": "quote-minimal attributed spans",
"locator": "Section 4, line 186",
"match_status": "exact-normalized-match",
"occurrence_ids": [
"V2-SOL-04:S3:S3-SP2"
],
"source_text_sha256": "014f247563e015dfe7cd6ea4dee6c410f57c74788a6fb5bd589f65df078e8c95",
"span_root_id": "span:sha256:3015a109e7898cb720e426e9a57c526277b648c75f1fb4b13631b00524eac943",
"text": "the cited URL does not exist"
},
{
"license_treatment": "quote-minimal attributed spans",
"locator": "Section 3.3, line 155",
"match_status": "exact-normalized-match",
"occurrence_ids": [
"V2-SOL-04:S3:S3-SP1"
],
"source_text_sha256": "014f247563e015dfe7cd6ea4dee6c410f57c74788a6fb5bd589f65df078e8c95",
"span_root_id": "span:sha256:6c1d327e58ef001533ea1c11010e974436bbeafbde138efcc7b516281f6d25a5",
"text": "78.87% vs. 72.94%"
},
{
"license_treatment": "quote-minimal attributed spans",
"locator": "p. 5, lines 315-329",
"match_status": "ordered-fragment-match",
"occurrence_ids": [
"V2-TERRA-01:s2_reportbench:s2_collection"
],
"source_text_sha256": "014f247563e015dfe7cd6ea4dee6c410f57c74788a6fb5bd589f65df078e8c95",
"span_root_id": "span:sha256:7a6046a756d6ec0315d330cf0817d062ea48498bdc9c90b3489d64b2bd12a3a4",
"text": "\"we manually collected responses ... during the period from July 14 to July 25... OpenAI was using the standard version of Deep Research, powered by the o3 model.\""
},
{
"license_treatment": "quote-minimal attributed spans",
"locator": "Section 2.2 'Cited statements'",
"match_status": "exact-normalized-match",
"occurrence_ids": [
"V2-TERRA-04:S3_reportbench:S3_method"
],
"source_text_sha256": "014f247563e015dfe7cd6ea4dee6c410f57c74788a6fb5bd589f65df078e8c95",
"span_root_id": "span:sha256:9c7a4c186cc85f185aa293a7c5a46c08f9dbbc1a9e63c449dc737447e42f8ee4",
"text": "\"retrieve the full content of each cited webpage\""
},
{
"license_treatment": "quote-minimal attributed spans",
"locator": "Settings and metrics, lines 132-139",
"match_status": "exact-normalized-match",
"occurrence_ids": [
"V2-TERRA-03:S3:S3-B"
],
"source_text_sha256": "014f247563e015dfe7cd6ea4dee6c410f57c74788a6fb5bd589f65df078e8c95",
"span_root_id": "span:sha256:ba6b8a8b7d9121ebb05d594176ceed3d1e9dbc637876783abfd0141ebbd0a822",
"text": "For statement extraction, supporting source extraction, and semantic consistency verification, we adopt gpt-4o."
},
{
"license_treatment": "quote-minimal attributed spans",
"locator": "Section 3.1 'Setttings'",
"match_status": "exact-normalized-match",
"occurrence_ids": [
"V2-TERRA-04:S3_reportbench:S3_time"
],
"source_text_sha256": "014f247563e015dfe7cd6ea4dee6c410f57c74788a6fb5bd589f65df078e8c95",
"span_root_id": "span:sha256:ddf4320a6d302c13d28b6225eb7e79ac15583349e1b449d4bb967eb89a257a3d",
"text": "\"during the period from July 14 to July 25\""
},
{
"license_treatment": "quote-minimal attributed spans",
"locator": "Table 1, OpenAI Deep Research row and cited-statements Match Rate column",
"match_status": "ordered-fragment-match",
"occurrence_ids": [
"V2-SOL-02:S5:S5_SPAN_1"
],
"source_text_sha256": "014f247563e015dfe7cd6ea4dee6c410f57c74788a6fb5bd589f65df078e8c95",
"span_root_id": "span:sha256:e223640cb0ed12c87ca1d2406f3276f30a3b8d2017dd4dc1457fe59a94471040",
"text": "OpenAI Deep Research ... Match Rate 78.87%"
}
],
"format": "epistemedia-quote-minimal-edition-projection-v0.1",
"license": "CC BY 4.0",
"license_treatment": "quote-minimal attributed spans",
"readback_receipts": [
{
"captured_bytes": 157386,
"captured_sha256": "055e568189a402dd510c7c84be60f59465d9a001808936bc25cb0026ac25d267",
"edition_id": "edition:reportbench-arxiv-v1",
"http_status": 200,
"media_type": "text/html; charset=utf-8",
"redirect_chain": [],
"requested_url": "https://arxiv.org/html/2508.15804",
"resolved_url": "https://arxiv.org/html/2508.15804",
"retrieval_status": "retrieved"
},
{
"captured_bytes": 157386,
"captured_sha256": "055e568189a402dd510c7c84be60f59465d9a001808936bc25cb0026ac25d267",
"edition_id": "edition:reportbench-arxiv-v1",
"http_status": 200,
"media_type": "text/html; charset=utf-8",
"redirect_chain": [],
"requested_url": "https://arxiv.org/html/2508.15804v1",
"resolved_url": "https://arxiv.org/html/2508.15804v1",
"retrieval_status": "retrieved"
},
{
"captured_bytes": 729607,
"captured_sha256": "90730ad75011d460305caf45a5be33b1b5eb4126f7e7efc1506f63beda1f91d3",
"edition_id": "edition:reportbench-arxiv-v1",
"http_status": 200,
"media_type": "application/pdf",
"redirect_chain": [],
"requested_url": "https://arxiv.org/pdf/2508.15804",
"resolved_url": "https://arxiv.org/pdf/2508.15804",
"retrieval_status": "retrieved"
}
],
"review_supplements": [],
"source_text_file": "reportbench.txt"
},
"content_digest": "sha256:e13193419b2e087faabf99c51b27acf1a58919072019e4197c41c279354eb2f8",
"content_length": 5426,
"edition_label": "Quote-minimal projection of edition:reportbench-arxiv-v1",
"id": "em:dossier-edition:sha256:c674e09fbcbadea10ca1f9c3bb17f5573f020fbc209410c85f3c8f56d53255b6",
"key": "edition-reportbench-arxiv-v1-25888dc44b",
"media_type": "application/json",
"retrieved_at": "2026-08-23T18:33:06Z",
"visibility": "public",
"work_key": "work-reportbench-dca823c910"
},
{
"content": {
"canonical_url": "https://arxiv.org/html/2507.16280v1",
"excerpts": [
{
"license_treatment": "metadata, digest, and quote-minimal attributed spans only",
"locator": "Section 4.2.1, Citation Support Verification and Score Computation",
"match_status": "exact-normalized-match",
"occurrence_ids": [
"V2-TERRA-02:s4_researcherbench:s4_method"
],
"source_text_sha256": "1aa77342c2571bb564fe1b3e9a2ba2333b4ef0d637c4a3ef622540741927b0ec",
"span_root_id": "span:sha256:886b50d94e3facabc487342cd2af5157cf956622ecfc368684596c5f6b61c350",
"text": "\"whether the extracted content supports the corresponding claim\""
},
{
"license_treatment": "metadata, digest, and quote-minimal attributed spans only",
"locator": "Section 5.1 and Appendix E, Table 4",
"match_status": "exact-normalized-match",
"occurrence_ids": [
"V2-TERRA-04:S2_researcherbench:S2_time"
],
"source_text_sha256": "1aa77342c2571bb564fe1b3e9a2ba2333b4ef0d637c4a3ef622540741927b0ec",
"span_root_id": "span:sha256:ea9cb2e1b0078c0857b789d61d0a0d26801cf6689186470588a87b98b4639796",
"text": "\"All evaluations were conducted between March and April in 2025\""
},
{
"license_treatment": "metadata, digest, and quote-minimal attributed spans only",
"locator": "Section 4.2.1 'Citation Support Verification' and Section 4.2 'Score Computation', equations (2)-(3)",
"match_status": "exact-normalized-match",
"occurrence_ids": [
"V2-TERRA-04:S2_researcherbench:S2_method"
],
"source_text_sha256": "1aa77342c2571bb564fe1b3e9a2ba2333b4ef0d637c4a3ef622540741927b0ec",
"span_root_id": "span:sha256:f96ad5daa81153a3689f52f5777fedfd70e8a87d44e7a1c9e2c22b5789a3707d",
"text": "\"whether the extracted content supports the corresponding claim\""
}
],
"format": "epistemedia-quote-minimal-edition-projection-v0.1",
"license": "arXiv non-exclusive distribution license",
"license_treatment": "metadata, digest, and quote-minimal attributed spans only",
"readback_receipts": [
{
"captured_bytes": 1566914,
"captured_sha256": "1571435270d3e781ab8d5254913af00a7dcf0efcb99ec8e6d3db87a7b5c92f02",
"edition_id": "edition:researcherbench-arxiv-v1",
"http_status": 200,
"media_type": "application/pdf",
"redirect_chain": [],
"requested_url": "https://arxiv.org/pdf/2507.16280",
"resolved_url": "https://arxiv.org/pdf/2507.16280",
"retrieval_status": "retrieved"
},
{
"captured_bytes": 280402,
"captured_sha256": "c080d7304274d70a39651f49297cc65910ca833d7ee6a3de3052370733e44ee3",
"edition_id": "edition:researcherbench-arxiv-v1",
"http_status": 200,
"media_type": "text/html; charset=utf-8",
"redirect_chain": [],
"requested_url": "https://arxiv.org/html/2507.16280",
"resolved_url": "https://arxiv.org/html/2507.16280",
"retrieval_status": "retrieved"
}
],
"review_supplements": [],
"source_text_file": "researcherbench.txt"
},
"content_digest": "sha256:dcf32e9c1c1a5f7a8517fc73c90e5d577ad4580919c45167ab12cba18d3db5dd",
"content_length": 2645,
"edition_label": "Quote-minimal projection of edition:researcherbench-arxiv-v1",
"id": "em:dossier-edition:sha256:e1cb847e48cc813a003705644839d51e54af0b84d06e0f8419d14a03764b93cd",
"key": "edition-researcherbench-arxiv-v1-39a8b256d7",
"media_type": "application/json",
"retrieved_at": "2026-08-23T18:33:06Z",
"visibility": "public",
"work_key": "work-researcherbench-82729ab60f"
},
{
"content": {
"canonical_url": "https://arxiv.org/html/2604.03173v1",
"excerpts": [
{
"license_treatment": "quote-minimal attributed spans",
"locator": "pp. 1-2, lines 64-95",
"match_status": "ordered-fragment-match",
"occurrence_ids": [
"V2-TERRA-01:s5_urlhealth:s5_scope"
],
"source_text_sha256": "d00cf613da17939694913654e404c5d61e78c62779fdd9d154655f5986fe643f",
"span_root_id": "span:sha256:55de6a6ca9414a252592e9f77544a9259533dd713ce1f12f6a216752a2b76e8f",
"text": "\"This study focuses on URL-based citation hallucinations\"; \"fabricated snippets ... and invented bibliographic entries ... require separate systematic study.\""
},
{
"license_treatment": "quote-minimal attributed spans",
"locator": "Section 2 definition discussion and Section 3.3",
"match_status": "exact-normalized-match",
"occurrence_ids": [
"V2-TERRA-04:S4_urlhealth:S4_limitations"
],
"source_text_sha256": "d00cf613da17939694913654e404c5d61e78c62779fdd9d154655f5986fe643f",
"span_root_id": "span:sha256:59d0da64f738e1d8745f2bff3663eb89e273f93fe501afe37e2494a8e0ff31a1",
"text": "\"our non-resolving rates are lower bounds\""
},
{
"license_treatment": "quote-minimal attributed spans",
"locator": "page 3, Section 3.3",
"match_status": "exact-normalized-match",
"occurrence_ids": [
"V2-SOL-01:s5_url_health:sp5a"
],
"source_text_sha256": "d00cf613da17939694913654e404c5d61e78c62779fdd9d154655f5986fe643f",
"span_root_id": "span:sha256:a18e778fac3d49e81167f05e09fbc361e91af8c0c01b0fec52991f0a1e3d0b91",
"text": "no archived snapshot exists at any timestamp"
},
{
"license_treatment": "quote-minimal attributed spans",
"locator": "Section 5.1 Results",
"match_status": "exact-normalized-match",
"occurrence_ids": [
"V2-SOL-03:s4:s4_span3"
],
"source_text_sha256": "d00cf613da17939694913654e404c5d61e78c62779fdd9d154655f5986fe643f",
"span_root_id": "span:sha256:b68153666e4e12aee8dc887fef0a61fa5f44b5f680a0b7eb95ff3defcf8b5bf1",
"text": "the non-resolving rate drops significantly for all three models"
},
{
"license_treatment": "quote-minimal attributed spans",
"locator": "Section 3.3 'URL extraction and classification'",
"match_status": "exact-normalized-match",
"occurrence_ids": [
"V2-TERRA-04:S4_urlhealth:S4_method"
],
"source_text_sha256": "d00cf613da17939694913654e404c5d61e78c62779fdd9d154655f5986fe643f",
"span_root_id": "span:sha256:bb62928b5e0cb4e373b2f6bfeb579d8564d05da5588f8b150a89ea4b9b2ed0db",
"text": "\"URLs returning 4xx or 5xx status codes, connection errors, or timeouts are classified as non-resolving\""
}
],
"format": "epistemedia-quote-minimal-edition-projection-v0.1",
"license": "CC0 1.0",
"license_treatment": "quote-minimal attributed spans",
"readback_receipts": [
{
"captured_bytes": 302498,
"captured_sha256": "332e5b5cb4b0ee7065b1bbc30436dfdfeafca8130fb87e32b0020026d8bbafe1",
"edition_id": "edition:url-health-arxiv-v1",
"http_status": 200,
"media_type": "text/html; charset=utf-8",
"redirect_chain": [],
"requested_url": "https://arxiv.org/html/2604.03173",
"resolved_url": "https://arxiv.org/html/2604.03173",
"retrieval_status": "retrieved"
},
{
"captured_bytes": 302498,
"captured_sha256": "332e5b5cb4b0ee7065b1bbc30436dfdfeafca8130fb87e32b0020026d8bbafe1",
"edition_id": "edition:url-health-arxiv-v1",
"http_status": 200,
"media_type": "text/html; charset=utf-8",
"redirect_chain": [],
"requested_url": "https://arxiv.org/html/2604.03173v1",
"resolved_url": "https://arxiv.org/html/2604.03173v1",
"retrieval_status": "retrieved"
},
{
"captured_bytes": 416944,
"captured_sha256": "e8145a9f62f2a2bf0c5f3e2607160c56e010d02da965b5f65dbfa8f07ddde06c",
"edition_id": "edition:url-health-arxiv-v1",
"http_status": 200,
"media_type": "application/pdf",
"redirect_chain": [],
"requested_url": "https://arxiv.org/pdf/2604.03173",
"resolved_url": "https://arxiv.org/pdf/2604.03173",
"retrieval_status": "retrieved"
},
{
"captured_bytes": 42408,
"captured_sha256": "3976893e82f91cc9e7826c0d2d08c6caab40778134dbab09ebbfe1f6f63f5397",
"edition_id": "edition:url-health-arxiv-v1",
"http_status": 200,
"media_type": "text/html; charset=utf-8",
"redirect_chain": [],
"requested_url": "https://arxiv.org/abs/2604.03173",
"resolved_url": "https://arxiv.org/abs/2604.03173",
"retrieval_status": "retrieved"
}
],
"review_supplements": [
{
"canonical_url": "https://arxiv.org/html/2604.03173v1",
"carrier": {
"bytes": 302498,
"media_type": "text/html; charset=utf-8",
"sha256": "332e5b5cb4b0ee7065b1bbc30436dfdfeafca8130fb87e32b0020026d8bbafe1"
},
"edition_id": "edition:url-health-arxiv-v1",
"extent": "DRBench (17) comprises 100 multilingual research queries (Chinese and English) covering finance, science, and technology, with pre-collected outputs from 23 models.",
"license_treatment": "quote-minimal attributed spans",
"locator": "Section 3 Experimental setup; Section 3.1 Datasets; section#S3.SS1; paragraph p#S3.SS1.p1.1",
"role": "dependence-qualification-only",
"supplement_id": "supplement:url-health-drbench-precollected-outputs",
"warrant_id": "warrant:url-health-resolution",
"work_id": "work:url-health"
}
],
"source_text_file": "urlhealth.txt"
},
"content_digest": "sha256:37bffd052d581c627c32d65958b6037bebdc9a9c4cf1bef8376499fb0caa5bb7",
"content_length": 4828,
"edition_label": "Quote-minimal projection of edition:url-health-arxiv-v1",
"id": "em:dossier-edition:sha256:89224c483a2c88c51351393bd2ae482b8f2642a27b621c5a7a1401b05fe86b2f",
"key": "edition-url-health-arxiv-v1-85fb934b53",
"media_type": "application/json",
"retrieved_at": "2026-08-23T18:33:06Z",
"visibility": "public",
"work_key": "work-url-health-f64342d493"
},
{
"content": {
"accepted_ledger_path": "research/how-we-know/agent-citation-lineage/evidence-ledger-v1.json",
"accepted_ledger_sha256": "29034edff99ae40f03d244998ee1d238917ac387422dea221b80897c94d556b3",
"accepted_review_path": "research/how-we-know/agent-citation-lineage/independent-review-receipt.json",
"accepted_review_sha256": "aadc83de7c4a57b60593124917bbb5fedd5e64c7aa50db3d71b4817626532ae4",
"accepted_source_readbacks_path": "research/how-we-know/agent-citation-lineage/source-readbacks-v1.json",
"accepted_source_readbacks_sha256": "58dd334d395fda27e04f2f692d2ad7278d12224fe5e03947b963671871609146",
"candidate_warrants": [
{
"canonical_proposition": "In one five-system dermatology-review audit, reference identifiability and metadata correctness did not establish sentence-level claim-citation concordance.",
"data_root": "data:keplinger-three-run-five-system-audit",
"derivation_root": "derivation:keplinger-published-tables-and-figure",
"independent_review_status": "pending",
"method_root": "method:human-reference-and-sentence-concordance-audit",
"review_status": "candidate-supported-with-scope",
"warrant_id": "warrant:keplinger-metadata-versus-support",
"work_ids": [
"work:keplinger-dermatology-audit",
"work:keplinger-supplement"
]
},
{
"canonical_proposition": "DeepResearch Bench evaluated binary support for deduplicated statement-URL pairs; numeric values are edition-specific.",
"data_root": "data:deepresearch-bench-100-tasks-and-system-outputs",
"derivation_root": "derivation:deepresearch-bench-edition-specific-tables",
"independent_review_status": "pending",
"method_root": "method:fact-statement-url-support-evaluator",
"review_status": "candidate-supported-with-edition-boundary",
"warrant_id": "warrant:deepresearch-bench-fact",
"work_ids": [
"work:deepresearch-bench-paper",
"work:deepresearch-bench-repository"
]
},
{
"canonical_proposition": "DeepTRACE measured statement-source support and found between-system variation; its Gemini value is internally inconsistent between table and prose.",
"data_root": "data:deeptrace-303-questions-2727-outputs",
"derivation_root": "derivation:deeptrace-tables-and-prose",
"independent_review_status": "pending",
"method_root": "method:deeptrace-statement-source-matrix",
"review_status": "candidate-supported-except-disputed-gemini-number",
"warrant_id": "warrant:deeptrace-support-variation",
"work_ids": [
"work:deeptrace"
]
},
{
"canonical_proposition": "Cited but Not Verified operationalized link access, topical relevance, and factual support separately and reported materially different rates.",
"data_root": "data:cnv-130-drbench-and-browsecomp-queries",
"derivation_root": "derivation:cnv-main-table",
"independent_review_status": "pending",
"method_root": "method:cnv-ast-access-relevance-fact-check",
"review_status": "candidate-supported-with-upstream-dependence",
"warrant_id": "warrant:cnv-link-relevance-support-gap",
"work_ids": [
"work:cited-not-verified"
]
},
{
"canonical_proposition": "Within the Cited but Not Verified harness, the 2-to-150-call ablation reduced reported Fact Check scores for two setups without establishing a general causal law about search depth.",
"data_root": "data:cnv-depth-ablation-two-model-setups",
"derivation_root": "derivation:cnv-depth-ablation-tables",
"independent_review_status": "pending",
"method_root": "method:cnv-ast-access-relevance-fact-check",
"review_status": "candidate-supported-as-within-harness-association",
"warrant_id": "warrant:cnv-search-depth-ablation",
"work_ids": [
"work:cited-not-verified"
]
},
{
"canonical_proposition": "The URL-health study measured HTTP resolution and Wayback presence, not semantic claim support, on outputs including reused DeepResearch Bench material.",
"data_root": "data:url-health-drbench-and-expertqa-urls",
"derivation_root": "derivation:url-health-observational-results",
"independent_review_status": "pending",
"method_root": "method:http-resolution-wayback-classification",
"review_status": "candidate-supported-with-semantic-exclusion",
"warrant_id": "warrant:url-health-resolution",
"work_ids": [
"work:url-health"
]
},
{
"canonical_proposition": "ReportBench compared cited statements with retrieved cited-page content and reported sub-100-percent semantic match rates in its bounded survey-task setting.",
"data_root": "data:reportbench-100-survey-tasks",
"derivation_root": "derivation:reportbench-table-1",
"independent_review_status": "pending",
"method_root": "method:reportbench-statement-source-semantic-match",
"review_status": "candidate-supported-with-judge-limitation",
"warrant_id": "warrant:reportbench-match-rate",
"work_ids": [
"work:reportbench"
]
}
],
"citation_occurrence_ids": [
"V2-SOL-01:s1_keplinger",
"V2-SOL-01:s1a_keplinger_data",
"V2-SOL-01:s2_drbench",
"V2-SOL-01:s2a_drbench_repo",
"V2-SOL-01:s3_deeptrace",
"V2-SOL-01:s4_cited_not_verified",
"V2-SOL-01:s5_url_health",
"V2-SOL-02:S1",
"V2-SOL-02:S2",
"V2-SOL-02:S3",
"V2-SOL-02:S4",
"V2-SOL-02:S5",
"V2-SOL-02:S6",
"V2-SOL-02:S7",
"V2-SOL-03:s1",
"V2-SOL-03:s2",
"V2-SOL-03:s3",
"V2-SOL-03:s4",
"V2-SOL-03:s5",
"V2-SOL-03:s6",
"V2-SOL-03:s7",
"V2-SOL-04:S1",
"V2-SOL-04:S2",
"V2-SOL-04:S3",
"V2-SOL-04:S4",
"V2-SOL-04:S5",
"V2-SOL-04:S6",
"V2-SOL-04:S7",
"V2-TERRA-01:s1_deepresearchbench",
"V2-TERRA-01:s2_reportbench",
"V2-TERRA-01:s3_researcherbench",
"V2-TERRA-01:s4_cited_not_verified",
"V2-TERRA-01:s5_urlhealth",
"V2-TERRA-02:s1_deepresearchbench",
"V2-TERRA-02:s2_liveresearchbench",
"V2-TERRA-02:s3_deeptrace",
"V2-TERRA-02:s4_researcherbench",
"V2-TERRA-03:S1",
"V2-TERRA-03:S2",
"V2-TERRA-03:S3",
"V2-TERRA-03:S4",
"V2-TERRA-03:S5",
"V2-TERRA-03:S6",
"V2-TERRA-04:S1_deepresearchbench",
"V2-TERRA-04:S2_researcherbench",
"V2-TERRA-04:S3_reportbench",
"V2-TERRA-04:S4_urlhealth",
"V2-TERRA-04:S5_cited_not_verified"
],
"cited_urls": [
"https://arxiv.org/abs/2604.03173",
"https://arxiv.org/abs/2605.06635",
"https://arxiv.org/abs/2607.08700",
"https://arxiv.org/html/2506.11763",
"https://arxiv.org/html/2506.11763v1",
"https://arxiv.org/html/2507.16280",
"https://arxiv.org/html/2508.15804",
"https://arxiv.org/html/2508.15804v1",
"https://arxiv.org/html/2509.04499",
"https://arxiv.org/html/2509.04499v1",
"https://arxiv.org/html/2604.03173",
"https://arxiv.org/html/2604.03173v1",
"https://arxiv.org/html/2605.06635",
"https://arxiv.org/html/2605.06635v1",
"https://arxiv.org/pdf/2507.16280",
"https://arxiv.org/pdf/2508.15804",
"https://arxiv.org/pdf/2510.14240v5",
"https://arxiv.org/pdf/2604.03173",
"https://arxiv.org/pdf/2605.06635",
"https://data.mendeley.com/datasets/3s73z9zf3c/1",
"https://data.mendeley.com/datasets/3s73z9zf3c/2",
"https://deepresearch-bench.github.io/static/papers/deepresearch-bench.pdf",
"https://github.com/Ayanami0730/deep_research_bench",
"https://onlinelibrary.wiley.com/doi/10.1111/jdv.70035",
"https://openreview.net/pdf?id=hQ0K2Hhq7H",
"https://pmc.ncbi.nlm.nih.gov/articles/PMC13109748/",
"https://proceedings.iclr.cc/paper_files/paper/2026/file/465f22be10e07b301c6ed58f0472f704-Paper-Conference.pdf",
"https://proceedings.iclr.cc/paper_files/paper/2026/file/ad08767706825033b99122332293033d-Paper-Conference.pdf",
"https://proceedings.iclr.cc/paper_files/paper/2026/hash/ad08767706825033b99122332293033d-Abstract-Conference.html",
"https://pubmed.ncbi.nlm.nih.gov/40904191/"
],
"corrections": [
{
"correction_id": "correction:liveresearchbench-license",
"effect": "license metadata corrected; scientific warrant unchanged",
"evidence": "The arXiv v5 abstract page links http://creativecommons.org/licenses/by-nc-sa/4.0/.",
"kind": "license",
"raw_occurrences": [
"V2-TERRA-02:s2_liveresearchbench"
],
"raw_value": "CC BY 4.0 for arXiv paper",
"review_value": "CC BY-NC-SA 4.0"
},
{
"correction_id": "correction:cnv-query-source-scope",
"effect": "raw scope receives no credit; canonical warrant retains both upstream roots",
"evidence": "Cited but Not Verified section 3.4 states both upstream query sources.",
"kind": "scope",
"raw_occurrences": [
"V2-TERRA-01:r4_cited_not_verified_source_attribution"
],
"raw_value": "130 research queries drawn from DeepResearch Bench",
"review_value": "130 research queries drawn from DeepResearch Bench and BrowseComp"
},
{
"correction_id": "correction:deeptrace-gemini-number",
"effect": "generic between-system variation remains; exact Gemini numeric warrant is withheld",
"evidence": "DeepTRACE ICLR 2026 PDF p.8 Table 1 and adjacent results prose disagree.",
"kind": "internal-source-conflict",
"raw_occurrences": [
"V2-SOL-01:r4_deeptrace_support",
"V2-SOL-03:r4_deeptrace_support",
"V2-SOL-04:R5",
"V2-TERRA-02:r4_deeptrace"
],
"raw_value": "Gemini Deep Research citation accuracy 50.3% from Table 1",
"review_value": "unresolved: Table 1 reports 50.3%, surrounding prose reports 40.3%"
},
{
"correction_id": "correction:drbench-edition-drift",
"effect": "method-level warrant remains; numeric roots are edition-specific and never pooled",
"evidence": "v1 reports 90.24/81.44/77.96 for Perplexity/Gemini/OpenAI; ICLR reports 82.63/78.30/75.01.",
"kind": "edition",
"raw_occurrences": [
"V2-SOL-01:r3_deepresearch_bench_fact",
"V2-SOL-02:R3_deepresearch_bench_fact",
"V2-SOL-04:R3",
"V2-TERRA-01:r1_deepresearchbench_fact",
"V2-TERRA-02:r1_deepresearchbench_fact",
"V2-TERRA-03:R3"
],
"raw_value": "arXiv v1 and ICLR values appear across the run matrix",
"review_value": "arXiv v1 and ICLR 2026 are separate editions with materially different FACT values"
},
{
"correction_id": "correction:four-independent-benchmarks",
"effect": "raw independence claim receives no credit",
"evidence": "The ledger records upstream DRBench reuse plus shared Jina Reader and judge-method roots.",
"kind": "dependence",
"raw_occurrences": [
"V2-TERRA-02:answer"
],
"raw_value": "Four independent benchmark papers",
"review_value": "separate papers with known shared task, retrieval, and LLM-judge dependencies"
},
{
"correction_id": "correction:mendeley-file-readback",
"effect": "landing-page metadata can support artifact identity; file-content spans receive no credit",
"evidence": "Fresh Mendeley landing-page reads returned 200; api.mendeley.com/datasets/3s73z9zf3c/versions/2 returned 401.",
"kind": "access",
"raw_occurrences": [
"V2-SOL-01:s1a_keplinger_data",
"V2-SOL-02:S3",
"V2-SOL-04:S6"
],
"raw_value": "supplementary files publicly retrieved or downloadable",
"review_value": "landing pages retrieved; credential-free API returned 401 and file bytes were not captured in this review"
},
{
"correction_id": "correction:independent-semantic-warrant-gaps",
"effect": "the complete raw proposition receives no credit; raw capture and narrower matched fragments remain visible",
"evidence": "Independent exact-edition review completed all 72 credited span roots and found these nine claim occurrences exceed the semantics of their linked spans.",
"kind": "semantic-warrant",
"raw_occurrences": [
"V2-SOL-02:R5_deeptrace_audit",
"V2-SOL-03:r3_deepresearch_bench_fact",
"V2-SOL-03:r8_search_depth_ablation",
"V2-SOL-03:r9_verifier_calibration",
"V2-SOL-04:R1",
"V2-SOL-04:R2",
"V2-SOL-04:R3",
"V2-SOL-04:R4",
"V2-SOL-04:R8"
],
"raw_value": "author-candidate support for the complete raw proposition",
"review_value": "linked quote fragments establish only a subset of the asserted method, comparison, metric, scope, or directional result"
}
],
"count_grammar": {
"candidate_warrant_root": "one reviewed proposition plus data, method, and derivation lineage with at least one claim whose matched spans semantically support the canonical proposition; it remains candidate until independent exact-head review",
"cited_url": "one distinct requested URL string in the eight raw answers",
"exact_span_root": "one edition, locator, and quote identity with an independently captured normalized or ordered-fragment match",
"examined_edition_root": "one versioned content state; mirrors and carriers do not create another edition",
"invalid_citation": "a malformed URL, a URL resolving to a different work, or a source identity contradicted by authoritative metadata",
"report": "one terminal answer artifact bound to one frozen v2 run slot",
"resolving_url_root": "one distinct requested URL whose fresh readback returned usable public source content",
"source_work_root": "one intellectual or data work after URL-carrier normalization",
"unresolved_citation": "one source occurrence with an inaccessible carrier, no independently matched span, or an unresolved identity or warrant correction"
},
"dependence_edges": [
{
"dimension": "upstream_citation",
"from": "work:cited-not-verified",
"note": "The paper states that its 130 queries are drawn from DeepResearch Bench and BrowseComp.",
"status": "known",
"to": "work:deepresearch-bench-paper"
},
{
"dimension": "upstream_citation",
"from": "work:url-health",
"note": "The URL-health study reuses DeepResearch Bench outputs for part of its corpus.",
"status": "known",
"to": "work:deepresearch-bench-paper"
},
{
"dimension": "data",
"from": "work:keplinger-supplement",
"note": "The supplement contains prompts, outputs, and evaluation material for the article.",
"status": "same-data-root",
"to": "work:keplinger-dermatology-audit"
},
{
"dimension": "data",
"from": "work:deepresearch-bench-repository",
"note": "The repository is the paper's official task/evaluator artifact, not independent evidence.",
"status": "same-data-root",
"to": "work:deepresearch-bench-paper"
},
{
"dimension": "method",
"from": "method:fact-statement-url-support-evaluator",
"status": "known",
"to": "method:llm-judge-after-web-extraction"
},
{
"dimension": "method",
"from": "method:deeptrace-statement-source-matrix",
"status": "known",
"to": "method:llm-judge-after-web-extraction"
},
{
"dimension": "method",
"from": "method:cnv-ast-access-relevance-fact-check",
"status": "known",
"to": "method:llm-judge-after-web-extraction"
},
{
"dimension": "method",
"from": "method:reportbench-statement-source-semantic-match",
"status": "known",
"to": "method:llm-judge-after-web-extraction"
},
{
"dimension": "method",
"from": "method:researcherbench-faithfulness-groundedness",
"status": "known",
"to": "method:llm-judge-after-web-extraction"
},
{
"dimension": "retrieval_infrastructure",
"from": "method:fact-statement-url-support-evaluator",
"status": "known",
"to": "retrieval:jina-reader"
},
{
"dimension": "retrieval_infrastructure",
"from": "method:deeptrace-statement-source-matrix",
"status": "known",
"to": "retrieval:jina-reader"
},
{
"dimension": "retrieval_infrastructure",
"from": "method:researcherbench-faithfulness-groundedness",
"status": "known",
"to": "retrieval:jina-reader"
},
{
"dimension": "data",
"from": "warrant:keplinger-metadata-versus-support",
"status": "declared-reviewed-lineage",
"to": "data:keplinger-three-run-five-system-audit"
},
{
"dimension": "method",
"from": "warrant:keplinger-metadata-versus-support",
"status": "declared-reviewed-lineage",
"to": "method:human-reference-and-sentence-concordance-audit"
},
{
"dimension": "derivation",
"from": "warrant:keplinger-metadata-versus-support",
"status": "declared-reviewed-lineage",
"to": "derivation:keplinger-published-tables-and-figure"
},
{
"dimension": "data",
"from": "warrant:deepresearch-bench-fact",
"status": "declared-reviewed-lineage",
"to": "data:deepresearch-bench-100-tasks-and-system-outputs"
},
{
"dimension": "method",
"from": "warrant:deepresearch-bench-fact",
"status": "declared-reviewed-lineage",
"to": "method:fact-statement-url-support-evaluator"
},
{
"dimension": "derivation",
"from": "warrant:deepresearch-bench-fact",
"status": "declared-reviewed-lineage",
"to": "derivation:deepresearch-bench-edition-specific-tables"
},
{
"dimension": "data",
"from": "warrant:deeptrace-support-variation",
"status": "declared-reviewed-lineage",
"to": "data:deeptrace-303-questions-2727-outputs"
},
{
"dimension": "method",
"from": "warrant:deeptrace-support-variation",
"status": "declared-reviewed-lineage",
"to": "method:deeptrace-statement-source-matrix"
},
{
"dimension": "derivation",
"from": "warrant:deeptrace-support-variation",
"status": "declared-reviewed-lineage",
"to": "derivation:deeptrace-tables-and-prose"
},
{
"dimension": "data",
"from": "warrant:cnv-link-relevance-support-gap",
"status": "declared-reviewed-lineage",
"to": "data:cnv-130-drbench-and-browsecomp-queries"
},
{
"dimension": "method",
"from": "warrant:cnv-link-relevance-support-gap",
"status": "declared-reviewed-lineage",
"to": "method:cnv-ast-access-relevance-fact-check"
},
{
"dimension": "derivation",
"from": "warrant:cnv-link-relevance-support-gap",
"status": "declared-reviewed-lineage",
"to": "derivation:cnv-main-table"
},
{
"dimension": "data",
"from": "warrant:cnv-search-depth-ablation",
"status": "declared-reviewed-lineage",
"to": "data:cnv-depth-ablation-two-model-setups"
},
{
"dimension": "method",
"from": "warrant:cnv-search-depth-ablation",
"status": "declared-reviewed-lineage",
"to": "method:cnv-ast-access-relevance-fact-check"
},
{
"dimension": "derivation",
"from": "warrant:cnv-search-depth-ablation",
"status": "declared-reviewed-lineage",
"to": "derivation:cnv-depth-ablation-tables"
},
{
"dimension": "data",
"from": "warrant:url-health-resolution",
"status": "declared-reviewed-lineage",
"to": "data:url-health-drbench-and-expertqa-urls"
},
{
"dimension": "method",
"from": "warrant:url-health-resolution",
"status": "declared-reviewed-lineage",
"to": "method:http-resolution-wayback-classification"
},
{
"dimension": "derivation",
"from": "warrant:url-health-resolution",
"status": "declared-reviewed-lineage",
"to": "derivation:url-health-observational-results"
},
{
"dimension": "data",
"from": "warrant:url-health-correction-loop",
"status": "declared-reviewed-lineage",
"to": "data:url-health-correction-experiment"
},
{
"dimension": "method",
"from": "warrant:url-health-correction-loop",
"status": "declared-reviewed-lineage",
"to": "method:url-checking-correction-loop"
},
{
"dimension": "derivation",
"from": "warrant:url-health-correction-loop",
"status": "declared-reviewed-lineage",
"to": "derivation:url-health-correction-results"
},
{
"dimension": "data",
"from": "warrant:reportbench-match-rate",
"status": "declared-reviewed-lineage",
"to": "data:reportbench-100-survey-tasks"
},
{
"dimension": "method",
"from": "warrant:reportbench-match-rate",
"status": "declared-reviewed-lineage",
"to": "method:reportbench-statement-source-semantic-match"
},
{
"dimension": "derivation",
"from": "warrant:reportbench-match-rate",
"status": "declared-reviewed-lineage",
"to": "derivation:reportbench-table-1"
},
{
"dimension": "data",
"from": "warrant:researcherbench-faithfulness-groundedness",
"status": "declared-reviewed-lineage",
"to": "data:researcherbench-65-frontier-ai-questions"
},
{
"dimension": "method",
"from": "warrant:researcherbench-faithfulness-groundedness",
"status": "declared-reviewed-lineage",
"to": "method:researcherbench-faithfulness-groundedness"
},
{
"dimension": "derivation",
"from": "warrant:researcherbench-faithfulness-groundedness",
"status": "declared-reviewed-lineage",
"to": "derivation:researcherbench-table-2"
},
{
"dimension": "data",
"from": "warrant:liveresearchbench-e1-e2-e3",
"status": "declared-reviewed-lineage",
"to": "data:liveresearchbench-wide-info-and-market-analysis"
},
{
"dimension": "method",
"from": "warrant:liveresearchbench-e1-e2-e3",
"status": "declared-reviewed-lineage",
"to": "method:liveresearchbench-e1-e2-e3-rubric-tree"
},
{
"dimension": "derivation",
"from": "warrant:liveresearchbench-e1-e2-e3",
"status": "declared-reviewed-lineage",
"to": "derivation:liveresearchbench-v5-table-7"
},
{
"dimension": "data",
"from": "warrant:citation-verifier-calibration",
"status": "declared-reviewed-lineage",
"to": "data:citationverbench-1248-human-reviewed-decisions"
},
{
"dimension": "method",
"from": "warrant:citation-verifier-calibration",
"status": "declared-reviewed-lineage",
"to": "method:citationverbench-rubric-judge-comparison"
},
{
"dimension": "derivation",
"from": "warrant:citation-verifier-calibration",
"status": "declared-reviewed-lineage",
"to": "derivation:citationverbench-reported-calibration"
},
{
"dimension": "model_profile",
"from": "report:V2-SOL-01",
"status": "requested-not-independent",
"to": "model-profile:gpt-5.6-sol"
},
{
"dimension": "prompt",
"from": "report:V2-SOL-01",
"status": "shared-exact-bytes",
"to": "prompt:sha256:d321a9cec7b5fe419157c0623e18ff0020cb0080fbe6dd9ed4fd20a0b896f670"
},
{
"dimension": "retrieval_infrastructure",
"from": "report:V2-SOL-01",
"status": "unknown-or-shared",
"to": "retrieval:codex-public-web-implementation-unknown"
},
{
"dimension": "exact_span",
"from": "citation:V2-SOL-01:s1_keplinger",
"status": "unresolved",
"to": "span:sha256:c29518951f6335c3c613ebed9c4babc2bc1903426992cd90118a999be69b5cdd"
},
{
"dimension": "exact_span",
"from": "citation:V2-SOL-01:s1_keplinger",
"status": "exact-normalized-match",
"to": "span:sha256:b903dc2284e1b4d80bb6ef948d79bbdbc7d7447e871281b8577284883f75ccfa"
},
{
"dimension": "exact_span",
"from": "citation:V2-SOL-01:s1_keplinger",
"status": "unresolved",
"to": "span:sha256:0bea3f00dfce1763dae86c14901bee21053f2d8ceee7f9dde4ab33ec78d3eb50"
},
{
"dimension": "requested_url",
"from": "citation:V2-SOL-01:s1_keplinger",
"status": "retrieved",
"to": "https://pmc.ncbi.nlm.nih.gov/articles/PMC13109748/"
},
{
"dimension": "source_work",
"from": "citation:V2-SOL-01:s1_keplinger",
"status": "normalized",
"to": "work:keplinger-dermatology-audit"
},
{
"dimension": "edition",
"from": "citation:V2-SOL-01:s1_keplinger",
"status": "examined",
"to": "edition:keplinger-vor-2025"
},
{
"dimension": "exact_span",
"from": "citation:V2-SOL-01:s1a_keplinger_data",
"status": "exact-normalized-match",
"to": "span:sha256:6f79c3c824ccf889d73e234335661e5fca68d90f8c78ee97aef488e4fdb32a49"
},
{
"dimension": "requested_url",
"from": "citation:V2-SOL-01:s1a_keplinger_data",
"status": "retrieved",
"to": "https://data.mendeley.com/datasets/3s73z9zf3c/2"
},
{
"dimension": "source_work",
"from": "citation:V2-SOL-01:s1a_keplinger_data",
"status": "normalized",
"to": "work:keplinger-supplement"
},
{
"dimension": "edition",
"from": "citation:V2-SOL-01:s1a_keplinger_data",
"status": "examined",
"to": "edition:keplinger-supplement-v2"
},
{
"dimension": "exact_span",
"from": "citation:V2-SOL-01:s2_drbench",
"status": "exact-normalized-match",
"to": "span:sha256:7d7331942f9d0520e18e96206e7718e5187788be4d48cef995fac79b235db757"
},
{
"dimension": "exact_span",
"from": "citation:V2-SOL-01:s2_drbench",
"status": "unresolved",
"to": "span:sha256:6070d8d9e7ba0d923aca7ae3517013c0e865119cea05c3e9cb0565d273eee0ee"
},
{
"dimension": "exact_span",
"from": "citation:V2-SOL-01:s2_drbench",
"status": "exact-normalized-match",
"to": "span:sha256:c27aaca02bda06ed764be53351158fc862af9d4a556d3e82074507436f48c3f8"
},
{
"dimension": "exact_span",
"from": "citation:V2-SOL-01:s2_drbench",
"status": "exact-normalized-match",
"to": "span:sha256:279db725380c133044944dccb379b75a396242219aa30437b26ddc40b09a3b5c"
},
{
"dimension": "requested_url",
"from": "citation:V2-SOL-01:s2_drbench",
"status": "retrieved",
"to": "https://proceedings.iclr.cc/paper_files/paper/2026/file/465f22be10e07b301c6ed58f0472f704-Paper-Conference.pdf"
},
{
"dimension": "source_work",
"from": "citation:V2-SOL-01:s2_drbench",
"status": "normalized",
"to": "work:deepresearch-bench-paper"
},
{
"dimension": "edition",
"from": "citation:V2-SOL-01:s2_drbench",
"status": "examined",
"to": "edition:drbench-iclr-2026"
},
{
"dimension": "exact_span",
"from": "citation:V2-SOL-01:s2a_drbench_repo",
"status": "exact-normalized-match",
"to": "span:sha256:0abbab3d6269d58eb76d8e90eff1e4c27512cf46d9b8e5ff0a9f537b18f232fa"
},
{
"dimension": "requested_url",
"from": "citation:V2-SOL-01:s2a_drbench_repo",
"status": "retrieved",
"to": "https://github.com/Ayanami0730/deep_research_bench"
},
{
"dimension": "source_work",
"from": "citation:V2-SOL-01:s2a_drbench_repo",
"status": "normalized",
"to": "work:deepresearch-bench-repository"
},
{
"dimension": "edition",
"from": "citation:V2-SOL-01:s2a_drbench_repo",
"status": "examined",
"to": "edition:drbench-repo-main-469cce5"
},
{
"dimension": "exact_span",
"from": "citation:V2-SOL-01:s3_deeptrace",
"status": "unresolved",
"to": "span:sha256:b45b782077327a7123cff72789773ad2f70f5608c8e0e33b0d211b3d60e0de0a"
},
{
"dimension": "exact_span",
"from": "citation:V2-SOL-01:s3_deeptrace",
"status": "unresolved",
"to": "span:sha256:4e55438723cddd7dfbd56c115e5fe6e809c0091ec7631928d67490aafc49f118"
},
{
"dimension": "exact_span",
"from": "citation:V2-SOL-01:s3_deeptrace",
"status": "exact-normalized-match",
"to": "span:sha256:a15ab02e9fcb23df02fb03d28967d4220ddf65ad380d223049f22b6ab68ed5eb"
},
{
"dimension": "exact_span",
"from": "citation:V2-SOL-01:s3_deeptrace",
"status": "exact-normalized-match",
"to": "span:sha256:35927851edd2e7e5e0f60d98498940f88304ba99bf1f85a08535663943c2f40d"
},
{
"dimension": "requested_url",
"from": "citation:V2-SOL-01:s3_deeptrace",
"status": "retrieved",
"to": "https://proceedings.iclr.cc/paper_files/paper/2026/file/ad08767706825033b99122332293033d-Paper-Conference.pdf"
},
{
"dimension": "source_work",
"from": "citation:V2-SOL-01:s3_deeptrace",
"status": "normalized",
"to": "work:deeptrace"
},
{
"dimension": "edition",
"from": "citation:V2-SOL-01:s3_deeptrace",
"status": "examined",
"to": "edition:deeptrace-iclr-2026"
},
{
"dimension": "exact_span",
"from": "citation:V2-SOL-01:s4_cited_not_verified",
"status": "unresolved",
"to": "span:sha256:9339db261929e3b56820214b791131dcbe342e20111e4b08916cfa523893810f"
},
{
"dimension": "exact_span",
"from": "citation:V2-SOL-01:s4_cited_not_verified",
"status": "unresolved",
"to": "span:sha256:419fcae33b7758515a2334e64b7dca713a139a11d6930bc9cc772559989ae12e"
},
{
"dimension": "exact_span",
"from": "citation:V2-SOL-01:s4_cited_not_verified",
"status": "exact-normalized-match",
"to": "span:sha256:c3dfc98566c27561eb7267972e00e3e1ba6a399a3ff5d002dca2d1233836790c"
},
{
"dimension": "requested_url",
"from": "citation:V2-SOL-01:s4_cited_not_verified",
"status": "retrieved",
"to": "https://arxiv.org/abs/2605.06635"
},
{
"dimension": "source_work",
"from": "citation:V2-SOL-01:s4_cited_not_verified",
"status": "normalized",
"to": "work:cited-not-verified"
},
{
"dimension": "edition",
"from": "citation:V2-SOL-01:s4_cited_not_verified",
"status": "examined",
"to": "edition:cnv-arxiv-v1"
},
{
"dimension": "exact_span",
"from": "citation:V2-SOL-01:s5_url_health",
"status": "exact-normalized-match",
"to": "span:sha256:a18e778fac3d49e81167f05e09fbc361e91af8c0c01b0fec52991f0a1e3d0b91"
},
{
"dimension": "exact_span",
"from": "citation:V2-SOL-01:s5_url_health",
"status": "unresolved",
"to": "span:sha256:db9dc28183620369f5ad1fb0f179e8aadbc14c69c0fdb57b8a8b79cd8e95e2a6"
},
{
"dimension": "exact_span",
"from": "citation:V2-SOL-01:s5_url_health",
"status": "unresolved",
"to": "span:sha256:541aadd42c71d1bd9c2c458c8e250d5000ffb75f3801eea832a0a0ef58a3914b"
},
{
"dimension": "requested_url",
"from": "citation:V2-SOL-01:s5_url_health",
"status": "retrieved",
"to": "https://arxiv.org/abs/2604.03173"
},
{
"dimension": "source_work",
"from": "citation:V2-SOL-01:s5_url_health",
"status": "normalized",
"to": "work:url-health"
},
{
"dimension": "edition",
"from": "citation:V2-SOL-01:s5_url_health",
"status": "examined",
"to": "edition:url-health-arxiv-v1"
},
{
"dimension": "upstream_citation",
"from": "claim:V2-SOL-01:r1_human_dermatology_chatgpt",
"status": "declared-but-citation-unresolved",
"to": "citation:V2-SOL-01:s1_keplinger"
},
{
"dimension": "upstream_citation",
"from": "claim:V2-SOL-01:r1_human_dermatology_chatgpt",
"status": "declared-but-citation-unresolved",
"to": "citation:V2-SOL-01:s1a_keplinger_data"
},
{
"dimension": "upstream_citation",
"from": "claim:V2-SOL-01:r2_human_dermatology_lechat",
"status": "declared-but-citation-unresolved",
"to": "citation:V2-SOL-01:s1_keplinger"
},
{
"dimension": "upstream_citation",
"from": "claim:V2-SOL-01:r2_human_dermatology_lechat",
"status": "declared-but-citation-unresolved",
"to": "citation:V2-SOL-01:s1a_keplinger_data"
},
{
"dimension": "upstream_citation",
"from": "claim:V2-SOL-01:r3_deepresearch_bench_fact",
"status": "declared-but-citation-unresolved",
"to": "citation:V2-SOL-01:s2_drbench"
},
{
"dimension": "upstream_citation",
"from": "claim:V2-SOL-01:r3_deepresearch_bench_fact",
"status": "declared-and-resolved",
"to": "citation:V2-SOL-01:s2a_drbench_repo"
},
{
"dimension": "upstream_citation",
"from": "claim:V2-SOL-01:r4_deeptrace_support",
"status": "declared-but-citation-unresolved",
"to": "citation:V2-SOL-01:s3_deeptrace"
},
{
"dimension": "upstream_citation",
"from": "claim:V2-SOL-01:r5_cited_but_not_verified_cross_model",
"status": "declared-but-citation-unresolved",
"to": "citation:V2-SOL-01:s4_cited_not_verified"
},
{
"dimension": "upstream_citation",
"from": "claim:V2-SOL-01:r6_cited_but_not_verified_depth_ablation",
"status": "declared-but-citation-unresolved",
"to": "citation:V2-SOL-01:s4_cited_not_verified"
},
{
"dimension": "upstream_citation",
"from": "claim:V2-SOL-01:r7_url_resolution_drbench",
"status": "declared-but-citation-unresolved",
"to": "citation:V2-SOL-01:s5_url_health"
},
{
"dimension": "model_profile",
"from": "report:V2-SOL-02",
"status": "requested-not-independent",
"to": "model-profile:gpt-5.6-sol"
},
{
"dimension": "prompt",
"from": "report:V2-SOL-02",
"status": "shared-exact-bytes",
"to": "prompt:sha256:d321a9cec7b5fe419157c0623e18ff0020cb0080fbe6dd9ed4fd20a0b896f670"
},
{
"dimension": "retrieval_infrastructure",
"from": "report:V2-SOL-02",
"status": "unknown-or-shared",
"to": "retrieval:codex-public-web-implementation-unknown"
},
{
"dimension": "exact_span",
"from": "citation:V2-SOL-02:S1",
"status": "unresolved",
"to": "span:sha256:a7e8b6b9e8aa380a3d1eaedaae1439792bc858a48735284ecbd2822da1c3c0fe"
},
{
"dimension": "exact_span",
"from": "citation:V2-SOL-02:S1",
"status": "unresolved",
"to": "span:sha256:388e78d0c793138d79699c82540cb28cbd5722adfe6df119b36ee40a4502d9b5"
},
{
"dimension": "requested_url",
"from": "citation:V2-SOL-02:S1",
"status": "retrieved",
"to": "https://arxiv.org/html/2604.03173v1"
},
{
"dimension": "source_work",
"from": "citation:V2-SOL-02:S1",
"status": "normalized",
"to": "work:url-health"
},
{
"dimension": "edition",
"from": "citation:V2-SOL-02:S1",
"status": "examined",
"to": "edition:url-health-arxiv-v1"
},
{
"dimension": "exact_span",
"from": "citation:V2-SOL-02:S2",
"status": "exact-normalized-match",
"to": "span:sha256:092745c7db6407d4d52927642741cde229fad05a5347bb1867a5154421060e70"
},
{
"dimension": "requested_url",
"from": "citation:V2-SOL-02:S2",
"status": "retrieved",
"to": "https://pmc.ncbi.nlm.nih.gov/articles/PMC13109748/"
},
{
"dimension": "source_work",
"from": "citation:V2-SOL-02:S2",
"status": "normalized",
"to": "work:keplinger-dermatology-audit"
},
{
"dimension": "edition",
"from": "citation:V2-SOL-02:S2",
"status": "examined",
"to": "edition:keplinger-vor-2025"
},
{
"dimension": "exact_span",
"from": "citation:V2-SOL-02:S3",
"status": "unresolved",
"to": "span:sha256:870d5c8e029e0c3d25b0a4682848e6caafc22ee35db116517a55bf334eea3e2d"
},
{
"dimension": "exact_span",
"from": "citation:V2-SOL-02:S3",
"status": "unresolved",
"to": "span:sha256:3bbe9293af4ece39817697cf29b0fe4ab9423d429d9867dec2540b72c82fd5ba"
},
{
"dimension": "requested_url",
"from": "citation:V2-SOL-02:S3",
"status": "retrieved",
"to": "https://data.mendeley.com/datasets/3s73z9zf3c/2"
},
{
"dimension": "source_work",
"from": "citation:V2-SOL-02:S3",
"status": "normalized",
"to": "work:keplinger-supplement"
},
{
"dimension": "edition",
"from": "citation:V2-SOL-02:S3",
"status": "examined",
"to": "edition:keplinger-supplement-v2"
},
{
"dimension": "exact_span",
"from": "citation:V2-SOL-02:S4",
"status": "unresolved",
"to": "span:sha256:c8683e7dad85f22158370acbb6a6e949b304b41bcba2c06f55cd31e90ce27130"
},
{
"dimension": "exact_span",
"from": "citation:V2-SOL-02:S4",
"status": "unresolved",
"to": "span:sha256:5899663b10cbe488c635c75743b9cf4ba1aab8c3e3a67604d4fa29248b387324"
},
{
"dimension": "requested_url",
"from": "citation:V2-SOL-02:S4",
"status": "inaccessible",
"to": "https://openreview.net/pdf?id=hQ0K2Hhq7H"
},
{
"dimension": "source_work",
"from": "citation:V2-SOL-02:S4",
"status": "normalized",
"to": "work:deepresearch-bench-paper"
},
{
"dimension": "edition",
"from": "citation:V2-SOL-02:S4",
"status": "examined",
"to": "edition:drbench-iclr-2026"
},
{
"dimension": "exact_span",
"from": "citation:V2-SOL-02:S5",
"status": "ordered-fragment-match",
"to": "span:sha256:e223640cb0ed12c87ca1d2406f3276f30a3b8d2017dd4dc1457fe59a94471040"
},
{
"dimension": "exact_span",
"from": "citation:V2-SOL-02:S5",
"status": "unresolved",
"to": "span:sha256:3b6f33fc8d800c8712c7b9fc2daafa5027f5176b23aa9120eeea323a21790a3a"
},
{
"dimension": "requested_url",
"from": "citation:V2-SOL-02:S5",
"status": "retrieved",
"to": "https://arxiv.org/html/2508.15804v1"
},
{
"dimension": "source_work",
"from": "citation:V2-SOL-02:S5",
"status": "normalized",
"to": "work:reportbench"
},
{
"dimension": "edition",
"from": "citation:V2-SOL-02:S5",
"status": "examined",
"to": "edition:reportbench-arxiv-v1"
},
{
"dimension": "exact_span",
"from": "citation:V2-SOL-02:S6",
"status": "exact-normalized-match",
"to": "span:sha256:a08468cb62f2feb21d9cbb95175ce00f54592bd83573cc9166e2375947f62fee"
},
{
"dimension": "exact_span",
"from": "citation:V2-SOL-02:S6",
"status": "exact-normalized-match",
"to": "span:sha256:e2c9f6220843a04665f5d1cd142ac15f04211ace5c579fa67f77e74e84a76dfe"
},
{
"dimension": "requested_url",
"from": "citation:V2-SOL-02:S6",
"status": "retrieved",
"to": "https://proceedings.iclr.cc/paper_files/paper/2026/hash/ad08767706825033b99122332293033d-Abstract-Conference.html"
},
{
"dimension": "source_work",
"from": "citation:V2-SOL-02:S6",
"status": "normalized",
"to": "work:deeptrace"
},
{
"dimension": "edition",
"from": "citation:V2-SOL-02:S6",
"status": "examined",
"to": "edition:deeptrace-iclr-2026"
},
{
"dimension": "exact_span",
"from": "citation:V2-SOL-02:S7",
"status": "exact-normalized-match",
"to": "span:sha256:912c931063109f84ea88aea34891dd5e0f9147d2176af258059246431e9abfb1"
},
{
"dimension": "exact_span",
"from": "citation:V2-SOL-02:S7",
"status": "unresolved",
"to": "span:sha256:1f0504defcd0c58184e6958c7cf30ddd2120357feb41c2753301a5912457fef9"
},
{
"dimension": "requested_url",
"from": "citation:V2-SOL-02:S7",
"status": "retrieved",
"to": "https://arxiv.org/html/2605.06635v1"
},
{
"dimension": "source_work",
"from": "citation:V2-SOL-02:S7",
"status": "normalized",
"to": "work:cited-not-verified"
},
{
"dimension": "edition",
"from": "citation:V2-SOL-02:S7",
"status": "examined",
"to": "edition:cnv-arxiv-v1"
},
{
"dimension": "upstream_citation",
"from": "claim:V2-SOL-02:R1_url_resolution",
"status": "declared-but-citation-unresolved",
"to": "citation:V2-SOL-02:S1"
},
{
"dimension": "upstream_citation",
"from": "claim:V2-SOL-02:R2_human_dermatology_claim_support",
"status": "declared-and-resolved",
"to": "citation:V2-SOL-02:S2"
},
{
"dimension": "upstream_citation",
"from": "claim:V2-SOL-02:R2_human_dermatology_claim_support",
"status": "declared-but-citation-unresolved",
"to": "citation:V2-SOL-02:S3"
},
{
"dimension": "upstream_citation",
"from": "claim:V2-SOL-02:R3_deepresearch_bench_fact",
"status": "declared-but-citation-unresolved",
"to": "citation:V2-SOL-02:S4"
},
{
"dimension": "upstream_citation",
"from": "claim:V2-SOL-02:R4_reportbench_match_rate",
"status": "declared-but-citation-unresolved",
"to": "citation:V2-SOL-02:S5"
},
{
"dimension": "upstream_citation",
"from": "claim:V2-SOL-02:R5_deeptrace_audit",
"status": "declared-and-resolved",
"to": "citation:V2-SOL-02:S6"
},
{
"dimension": "upstream_citation",
"from": "claim:V2-SOL-02:R6_surface_quality_vs_fact_support",
"status": "declared-but-citation-unresolved",
"to": "citation:V2-SOL-02:S7"
},
{
"dimension": "upstream_citation",
"from": "claim:V2-SOL-02:R7_search_depth_ablation",
"status": "declared-but-citation-unresolved",
"to": "citation:V2-SOL-02:S7"
},
{
"dimension": "model_profile",
"from": "report:V2-SOL-03",
"status": "requested-not-independent",
"to": "model-profile:gpt-5.6-sol"
},
{
"dimension": "prompt",
"from": "report:V2-SOL-03",
"status": "shared-exact-bytes",
"to": "prompt:sha256:d321a9cec7b5fe419157c0623e18ff0020cb0080fbe6dd9ed4fd20a0b896f670"
},
{
"dimension": "retrieval_infrastructure",
"from": "report:V2-SOL-03",
"status": "unknown-or-shared",
"to": "retrieval:codex-public-web-implementation-unknown"
},
{
"dimension": "exact_span",
"from": "citation:V2-SOL-03:s1",
"status": "exact-normalized-match",
"to": "span:sha256:46c392bc942cd88d525d74298e6bc6ebd9aeaec533ddfb7b7b761a9f5b454b05"
},
{
"dimension": "requested_url",
"from": "citation:V2-SOL-03:s1",
"status": "inaccessible",
"to": "https://onlinelibrary.wiley.com/doi/10.1111/jdv.70035"
},
{
"dimension": "source_work",
"from": "citation:V2-SOL-03:s1",
"status": "normalized",
"to": "work:keplinger-dermatology-audit"
},
{
"dimension": "edition",
"from": "citation:V2-SOL-03:s1",
"status": "examined",
"to": "edition:keplinger-vor-2025"
},
{
"dimension": "exact_span",
"from": "citation:V2-SOL-03:s2",
"status": "exact-normalized-match",
"to": "span:sha256:2f311bfb451c6208101cb7583af37c338914fcdb10f5c75b131e23d5bb08e66a"
},
{
"dimension": "requested_url",
"from": "citation:V2-SOL-03:s2",
"status": "retrieved",
"to": "https://arxiv.org/html/2506.11763"
},
{
"dimension": "source_work",
"from": "citation:V2-SOL-03:s2",
"status": "normalized",
"to": "work:deepresearch-bench-paper"
},
{
"dimension": "edition",
"from": "citation:V2-SOL-03:s2",
"status": "examined",
"to": "edition:drbench-arxiv-v1"
},
{
"dimension": "exact_span",
"from": "citation:V2-SOL-03:s3",
"status": "unresolved",
"to": "span:sha256:91dd7d34a9546d71c8789056ebde06de25247281b1fb03ff41fc3a3ff20e9520"
},
{
"dimension": "requested_url",
"from": "citation:V2-SOL-03:s3",
"status": "retrieved",
"to": "https://arxiv.org/html/2509.04499"
},
{
"dimension": "source_work",
"from": "citation:V2-SOL-03:s3",
"status": "normalized",
"to": "work:deeptrace"
},
{
"dimension": "edition",
"from": "citation:V2-SOL-03:s3",
"status": "examined",
"to": "edition:deeptrace-arxiv-v1"
},
{
"dimension": "exact_span",
"from": "citation:V2-SOL-03:s4",
"status": "unresolved",
"to": "span:sha256:c130950377ad7bcb4218863dbbb72f86c43005d543056438e07501f32d369df7"
},
{
"dimension": "exact_span",
"from": "citation:V2-SOL-03:s4",
"status": "unresolved",
"to": "span:sha256:814248358095099e8d04475d7f17ce5aff8b4fe64f11239ea48edaf4d9a4b030"
},
{
"dimension": "exact_span",
"from": "citation:V2-SOL-03:s4",
"status": "exact-normalized-match",
"to": "span:sha256:b68153666e4e12aee8dc887fef0a61fa5f44b5f680a0b7eb95ff3defcf8b5bf1"
},
{
"dimension": "requested_url",
"from": "citation:V2-SOL-03:s4",
"status": "retrieved",
"to": "https://arxiv.org/html/2604.03173"
},
{
"dimension": "source_work",
"from": "citation:V2-SOL-03:s4",
"status": "normalized",
"to": "work:url-health"
},
{
"dimension": "edition",
"from": "citation:V2-SOL-03:s4",
"status": "examined",
"to": "edition:url-health-arxiv-v1"
},
{
"dimension": "exact_span",
"from": "citation:V2-SOL-03:s5",
"status": "exact-normalized-match",
"to": "span:sha256:bb16d5f06abe8641d9ea6694e549aa15e240dd88946aa47a2d4e30ddd3351a24"
},
{
"dimension": "exact_span",
"from": "citation:V2-SOL-03:s5",
"status": "exact-normalized-match",
"to": "span:sha256:6e961498ed0f81a950f775956587f741e0eb9f11640fd3cb40870585558d869d"
},
{
"dimension": "requested_url",
"from": "citation:V2-SOL-03:s5",
"status": "retrieved",
"to": "https://arxiv.org/html/2605.06635"
},
{
"dimension": "source_work",
"from": "citation:V2-SOL-03:s5",
"status": "normalized",
"to": "work:cited-not-verified"
},
{
"dimension": "edition",
"from": "citation:V2-SOL-03:s5",
"status": "examined",
"to": "edition:cnv-arxiv-v1"
},
{
"dimension": "exact_span",
"from": "citation:V2-SOL-03:s6",
"status": "exact-normalized-match",
"to": "span:sha256:6885cf2524462022475b5829da081b2d35fcfe7463ca6ad0979e7c4fb1f0b54d"
},
{
"dimension": "requested_url",
"from": "citation:V2-SOL-03:s6",
"status": "retrieved",
"to": "https://arxiv.org/abs/2607.08700"
},
{
"dimension": "source_work",
"from": "citation:V2-SOL-03:s6",
"status": "normalized",
"to": "work:citation-verifier-benchmark"
},
{
"dimension": "edition",
"from": "citation:V2-SOL-03:s6",
"status": "examined",
"to": "edition:citation-verifier-arxiv-v1"
},
{
"dimension": "exact_span",
"from": "citation:V2-SOL-03:s7",
"status": "exact-normalized-match",
"to": "span:sha256:e5bcf0241a1aeae7d9a92755b1bcf0adeeea93a8922f22f832693c61780386e7"
},
{
"dimension": "requested_url",
"from": "citation:V2-SOL-03:s7",
"status": "retrieved",
"to": "https://github.com/Ayanami0730/deep_research_bench"
},
{
"dimension": "source_work",
"from": "citation:V2-SOL-03:s7",
"status": "normalized",
"to": "work:deepresearch-bench-repository"
},
{
"dimension": "edition",
"from": "citation:V2-SOL-03:s7",
"status": "examined",
"to": "edition:drbench-repo-main-469cce5"
},
{
"dimension": "upstream_citation",
"from": "claim:V2-SOL-03:r1_reference_metadata_human_audit",
"status": "declared-but-citation-unresolved",
"to": "citation:V2-SOL-03:s1"
},
{
"dimension": "upstream_citation",
"from": "claim:V2-SOL-03:r2_claim_support_human_audit",
"status": "declared-but-citation-unresolved",
"to": "citation:V2-SOL-03:s1"
},
{
"dimension": "upstream_citation",
"from": "claim:V2-SOL-03:r3_deepresearch_bench_fact",
"status": "declared-and-resolved",
"to": "citation:V2-SOL-03:s2"
},
{
"dimension": "upstream_citation",
"from": "claim:V2-SOL-03:r3_deepresearch_bench_fact",
"status": "declared-and-resolved",
"to": "citation:V2-SOL-03:s7"
},
{
"dimension": "upstream_citation",
"from": "claim:V2-SOL-03:r4_deeptrace_support",
"status": "declared-but-citation-unresolved",
"to": "citation:V2-SOL-03:s3"
},
{
"dimension": "upstream_citation",
"from": "claim:V2-SOL-03:r5_deep_agent_url_resolution",
"status": "declared-but-citation-unresolved",
"to": "citation:V2-SOL-03:s4"
},
{
"dimension": "upstream_citation",
"from": "claim:V2-SOL-03:r5_deep_agent_url_resolution",
"status": "declared-and-resolved",
"to": "citation:V2-SOL-03:s2"
},
{
"dimension": "upstream_citation",
"from": "claim:V2-SOL-03:r6_url_self_correction",
"status": "declared-but-citation-unresolved",
"to": "citation:V2-SOL-03:s4"
},
{
"dimension": "upstream_citation",
"from": "claim:V2-SOL-03:r7_source_attribution_framework",
"status": "declared-and-resolved",
"to": "citation:V2-SOL-03:s5"
},
{
"dimension": "upstream_citation",
"from": "claim:V2-SOL-03:r8_search_depth_ablation",
"status": "declared-and-resolved",
"to": "citation:V2-SOL-03:s5"
},
{
"dimension": "upstream_citation",
"from": "claim:V2-SOL-03:r9_verifier_calibration",
"status": "declared-and-resolved",
"to": "citation:V2-SOL-03:s6"
},
{
"dimension": "model_profile",
"from": "report:V2-SOL-04",
"status": "requested-not-independent",
"to": "model-profile:gpt-5.6-sol"
},
{
"dimension": "prompt",
"from": "report:V2-SOL-04",
"status": "shared-exact-bytes",
"to": "prompt:sha256:d321a9cec7b5fe419157c0623e18ff0020cb0080fbe6dd9ed4fd20a0b896f670"
},
{
"dimension": "retrieval_infrastructure",
"from": "report:V2-SOL-04",
"status": "unknown-or-shared",
"to": "retrieval:codex-public-web-implementation-unknown"
},
{
"dimension": "exact_span",
"from": "citation:V2-SOL-04:S1",
"status": "exact-normalized-match",
"to": "span:sha256:74f6acafbb4c54d556a572460bf1b45bacd1e293f4aa16cc2d499174096b9c20"
},
{
"dimension": "exact_span",
"from": "citation:V2-SOL-04:S1",
"status": "exact-normalized-match",
"to": "span:sha256:deada8bacb987a47f2a42e52fbfeb5c1beb3d250c0be79e2f9c5cdd6e16b4bc7"
},
{
"dimension": "exact_span",
"from": "citation:V2-SOL-04:S1",
"status": "exact-normalized-match",
"to": "span:sha256:417182b4c15ac550469e435a02aab1e91b395f72f186312de9d34d57fd1c6c98"
},
{
"dimension": "requested_url",
"from": "citation:V2-SOL-04:S1",
"status": "retrieved",
"to": "https://arxiv.org/html/2605.06635v1"
},
{
"dimension": "source_work",
"from": "citation:V2-SOL-04:S1",
"status": "normalized",
"to": "work:cited-not-verified"
},
{
"dimension": "edition",
"from": "citation:V2-SOL-04:S1",
"status": "examined",
"to": "edition:cnv-arxiv-v1"
},
{
"dimension": "exact_span",
"from": "citation:V2-SOL-04:S2",
"status": "exact-normalized-match",
"to": "span:sha256:196c54d6212fbc0c054dbf8e8467f2388c1f788bb93f4b1a4b6ae085eed17685"
},
{
"dimension": "exact_span",
"from": "citation:V2-SOL-04:S2",
"status": "exact-normalized-match",
"to": "span:sha256:13cb0ddb087b82988b7445a9cfb02562f9c044ef4f244e8b176ff475859eaea3"
},
{
"dimension": "exact_span",
"from": "citation:V2-SOL-04:S2",
"status": "exact-normalized-match",
"to": "span:sha256:b83996e3b5fdf878e04d6d41d0e7a1eebee9fdd9bbccf2ae4eaf33524fcfca44"
},
{
"dimension": "exact_span",
"from": "citation:V2-SOL-04:S2",
"status": "exact-normalized-match",
"to": "span:sha256:f35c9565f2fd047f6e7272dc608bbf6a63813a717bad5354c24bdaac65e40bfe"
},
{
"dimension": "requested_url",
"from": "citation:V2-SOL-04:S2",
"status": "retrieved",
"to": "https://arxiv.org/html/2506.11763v1"
},
{
"dimension": "source_work",
"from": "citation:V2-SOL-04:S2",
"status": "normalized",
"to": "work:deepresearch-bench-paper"
},
{
"dimension": "edition",
"from": "citation:V2-SOL-04:S2",
"status": "examined",
"to": "edition:drbench-arxiv-v1"
},
{
"dimension": "exact_span",
"from": "citation:V2-SOL-04:S3",
"status": "exact-normalized-match",
"to": "span:sha256:6c1d327e58ef001533ea1c11010e974436bbeafbde138efcc7b516281f6d25a5"
},
{
"dimension": "exact_span",
"from": "citation:V2-SOL-04:S3",
"status": "exact-normalized-match",
"to": "span:sha256:3015a109e7898cb720e426e9a57c526277b648c75f1fb4b13631b00524eac943"
},
{
"dimension": "requested_url",
"from": "citation:V2-SOL-04:S3",
"status": "retrieved",
"to": "https://arxiv.org/html/2508.15804v1"
},
{
"dimension": "source_work",
"from": "citation:V2-SOL-04:S3",
"status": "normalized",
"to": "work:reportbench"
},
{
"dimension": "edition",
"from": "citation:V2-SOL-04:S3",
"status": "examined",
"to": "edition:reportbench-arxiv-v1"
},
{
"dimension": "exact_span",
"from": "citation:V2-SOL-04:S4",
"status": "exact-normalized-match",
"to": "span:sha256:8cce3d9ee1e1674ba1b93321dd9b99a55d834ca980451f8ab61bcaf5bdda8e4d"
},
{
"dimension": "exact_span",
"from": "citation:V2-SOL-04:S4",
"status": "exact-normalized-match",
"to": "span:sha256:9dd6eb46529e77a62748093528e52685fbd92d46a756aef750d22aa25981db23"
},
{
"dimension": "requested_url",
"from": "citation:V2-SOL-04:S4",
"status": "retrieved",
"to": "https://arxiv.org/html/2509.04499v1"
},
{
"dimension": "source_work",
"from": "citation:V2-SOL-04:S4",
"status": "normalized",
"to": "work:deeptrace"
},
{
"dimension": "edition",
"from": "citation:V2-SOL-04:S4",
"status": "examined",
"to": "edition:deeptrace-arxiv-v1"
},
{
"dimension": "exact_span",
"from": "citation:V2-SOL-04:S5",
"status": "exact-normalized-match",
"to": "span:sha256:8c8d9a6dc003bf0298542b85df3f6f08db631e68e89f606f4e53e36e86dde167"
},
{
"dimension": "exact_span",
"from": "citation:V2-SOL-04:S5",
"status": "exact-normalized-match",
"to": "span:sha256:39a9931f05ea356fe904ed15768fa08c66440b1fcb20f98e267a315e6d562af9"
},
{
"dimension": "exact_span",
"from": "citation:V2-SOL-04:S5",
"status": "exact-normalized-match",
"to": "span:sha256:f05b3189dc3fafd45b5bde643a91fd0616832928215a37a2430a5a658e592029"
},
{
"dimension": "requested_url",
"from": "citation:V2-SOL-04:S5",
"status": "retrieved",
"to": "https://pmc.ncbi.nlm.nih.gov/articles/PMC13109748/"
},
{
"dimension": "source_work",
"from": "citation:V2-SOL-04:S5",
"status": "normalized",
"to": "work:keplinger-dermatology-audit"
},
{
"dimension": "edition",
"from": "citation:V2-SOL-04:S5",
"status": "examined",
"to": "edition:keplinger-vor-2025"
},
{
"dimension": "exact_span",
"from": "citation:V2-SOL-04:S6",
"status": "exact-normalized-match",
"to": "span:sha256:94ad583514aa7e04b8d5f7b8f9cdda6fdffe8ca6712f033da3ce4f18fe7c03f1"
},
{
"dimension": "requested_url",
"from": "citation:V2-SOL-04:S6",
"status": "retrieved",
"to": "https://data.mendeley.com/datasets/3s73z9zf3c/1"
},
{
"dimension": "source_work",
"from": "citation:V2-SOL-04:S6",
"status": "normalized",
"to": "work:keplinger-supplement"
},
{
"dimension": "edition",
"from": "citation:V2-SOL-04:S6",
"status": "examined",
"to": "edition:keplinger-supplement-v1"
},
{
"dimension": "exact_span",
"from": "citation:V2-SOL-04:S7",
"status": "exact-normalized-match",
"to": "span:sha256:e54ee1457854a9a5e46f7c7c2e66ac6a5a69b963f8ee3450f40052f138d5d0c1"
},
{
"dimension": "exact_span",
"from": "citation:V2-SOL-04:S7",
"status": "exact-normalized-match",
"to": "span:sha256:794bdf7e0595799afbb0d55f3dc522077aed9615c59f1e3575707148cf444bcf"
},
{
"dimension": "requested_url",
"from": "citation:V2-SOL-04:S7",
"status": "retrieved",
"to": "https://arxiv.org/abs/2607.08700"
},
{
"dimension": "source_work",
"from": "citation:V2-SOL-04:S7",
"status": "normalized",
"to": "work:citation-verifier-benchmark"
},
{
"dimension": "edition",
"from": "citation:V2-SOL-04:S7",
"status": "examined",
"to": "edition:citation-verifier-arxiv-v1"
},
{
"dimension": "upstream_citation",
"from": "claim:V2-SOL-04:R1",
"status": "declared-and-resolved",
"to": "citation:V2-SOL-04:S1"
},
{
"dimension": "upstream_citation",
"from": "claim:V2-SOL-04:R2",
"status": "declared-and-resolved",
"to": "citation:V2-SOL-04:S1"
},
{
"dimension": "upstream_citation",
"from": "claim:V2-SOL-04:R3",
"status": "declared-and-resolved",
"to": "citation:V2-SOL-04:S2"
},
{
"dimension": "upstream_citation",
"from": "claim:V2-SOL-04:R4",
"status": "declared-and-resolved",
"to": "citation:V2-SOL-04:S3"
},
{
"dimension": "upstream_citation",
"from": "claim:V2-SOL-04:R5",
"status": "declared-and-resolved",
"to": "citation:V2-SOL-04:S4"
},
{
"dimension": "upstream_citation",
"from": "claim:V2-SOL-04:R6",
"status": "declared-and-resolved",
"to": "citation:V2-SOL-04:S5"
},
{
"dimension": "upstream_citation",
"from": "claim:V2-SOL-04:R6",
"status": "declared-but-citation-unresolved",
"to": "citation:V2-SOL-04:S6"
},
{
"dimension": "upstream_citation",
"from": "claim:V2-SOL-04:R7",
"status": "declared-and-resolved",
"to": "citation:V2-SOL-04:S5"
},
{
"dimension": "upstream_citation",
"from": "claim:V2-SOL-04:R7",
"status": "declared-but-citation-unresolved",
"to": "citation:V2-SOL-04:S6"
},
{
"dimension": "upstream_citation",
"from": "claim:V2-SOL-04:R8",
"status": "declared-and-resolved",
"to": "citation:V2-SOL-04:S7"
},
{
"dimension": "model_profile",
"from": "report:V2-TERRA-01",
"status": "requested-not-independent",
"to": "model-profile:gpt-5.6-terra"
},
{
"dimension": "prompt",
"from": "report:V2-TERRA-01",
"status": "shared-exact-bytes",
"to": "prompt:sha256:d321a9cec7b5fe419157c0623e18ff0020cb0080fbe6dd9ed4fd20a0b896f670"
},
{
"dimension": "retrieval_infrastructure",
"from": "report:V2-TERRA-01",
"status": "unknown-or-shared",
"to": "retrieval:codex-public-web-implementation-unknown"
},
{
"dimension": "exact_span",
"from": "citation:V2-TERRA-01:s1_deepresearchbench",
"status": "ordered-fragment-match",
"to": "span:sha256:b0a39c963cdfdc5f9852cfc9f7e33e05d31232bd1d8998851e879c38e6c34843"
},
{
"dimension": "exact_span",
"from": "citation:V2-TERRA-01:s1_deepresearchbench",
"status": "unresolved",
"to": "span:sha256:77001af3bf7a976754d0b5d149da1c4e9b464c3515694958fe01a3c4ca8c7a5c"
},
{
"dimension": "exact_span",
"from": "citation:V2-TERRA-01:s1_deepresearchbench",
"status": "unresolved",
"to": "span:sha256:34a61ece593714d9608073ee5aeea22635a3fb111582fc8e2e8bf0b371955b70"
},
{
"dimension": "requested_url",
"from": "citation:V2-TERRA-01:s1_deepresearchbench",
"status": "retrieved",
"to": "https://deepresearch-bench.github.io/static/papers/deepresearch-bench.pdf"
},
{
"dimension": "source_work",
"from": "citation:V2-TERRA-01:s1_deepresearchbench",
"status": "normalized",
"to": "work:deepresearch-bench-paper"
},
{
"dimension": "edition",
"from": "citation:V2-TERRA-01:s1_deepresearchbench",
"status": "examined",
"to": "edition:drbench-arxiv-v1"
},
{
"dimension": "exact_span",
"from": "citation:V2-TERRA-01:s2_reportbench",
"status": "unresolved",
"to": "span:sha256:f4c499ea114f2647a0dc4830b72e4b9120d6c368f846efb1ccd3ae4c0ed4f2c7"
},
{
"dimension": "exact_span",
"from": "citation:V2-TERRA-01:s2_reportbench",
"status": "unresolved",
"to": "span:sha256:a6bde87f202e383e8bdd8403b2e5241b65380e0616e61b18f36a7c3719b55d76"
},
{
"dimension": "exact_span",
"from": "citation:V2-TERRA-01:s2_reportbench",
"status": "ordered-fragment-match",
"to": "span:sha256:7a6046a756d6ec0315d330cf0817d062ea48498bdc9c90b3489d64b2bd12a3a4"
},
{
"dimension": "requested_url",
"from": "citation:V2-TERRA-01:s2_reportbench",
"status": "retrieved",
"to": "https://arxiv.org/pdf/2508.15804"
},
{
"dimension": "source_work",
"from": "citation:V2-TERRA-01:s2_reportbench",
"status": "normalized",
"to": "work:reportbench"
},
{
"dimension": "edition",
"from": "citation:V2-TERRA-01:s2_reportbench",
"status": "examined",
"to": "edition:reportbench-arxiv-v1"
},
{
"dimension": "exact_span",
"from": "citation:V2-TERRA-01:s3_researcherbench",
"status": "unresolved",
"to": "span:sha256:5351ba5f0c5afba376e1a70ffafab3d6a4eefae1d9507b5206a64951f360a095"
},
{
"dimension": "exact_span",
"from": "citation:V2-TERRA-01:s3_researcherbench",
"status": "unresolved",
"to": "span:sha256:5fb35a076c774fd7141fd642f5490f3679fc2452e9e886b62d2e3c370b9c029d"
},
{
"dimension": "exact_span",
"from": "citation:V2-TERRA-01:s3_researcherbench",
"status": "unresolved",
"to": "span:sha256:eadcf02c45f4b152c73c6f871d116e7d1db8e78c796671580a5eb72e7ff16687"
},
{
"dimension": "requested_url",
"from": "citation:V2-TERRA-01:s3_researcherbench",
"status": "retrieved",
"to": "https://arxiv.org/pdf/2507.16280"
},
{
"dimension": "source_work",
"from": "citation:V2-TERRA-01:s3_researcherbench",
"status": "normalized",
"to": "work:researcherbench"
},
{
"dimension": "edition",
"from": "citation:V2-TERRA-01:s3_researcherbench",
"status": "examined",
"to": "edition:researcherbench-arxiv-v1"
},
{
"dimension": "exact_span",
"from": "citation:V2-TERRA-01:s4_cited_not_verified",
"status": "unresolved",
"to": "span:sha256:48a692e799a44e37783d312cb80a0500cca0aa94abb30dc074c1cb6f0bb0ead1"
},
{
"dimension": "exact_span",
"from": "citation:V2-TERRA-01:s4_cited_not_verified",
"status": "unresolved",
"to": "span:sha256:a9625285d829a0463b566e89d328e49b9c21cc07e83dbf845ac422f5a8e4b72e"
},
{
"dimension": "requested_url",
"from": "citation:V2-TERRA-01:s4_cited_not_verified",
"status": "retrieved",
"to": "https://arxiv.org/pdf/2605.06635"
},
{
"dimension": "source_work",
"from": "citation:V2-TERRA-01:s4_cited_not_verified",
"status": "normalized",
"to": "work:cited-not-verified"
},
{
"dimension": "edition",
"from": "citation:V2-TERRA-01:s4_cited_not_verified",
"status": "examined",
"to": "edition:cnv-arxiv-v1"
},
{
"dimension": "exact_span",
"from": "citation:V2-TERRA-01:s5_urlhealth",
"status": "ordered-fragment-match",
"to": "span:sha256:55de6a6ca9414a252592e9f77544a9259533dd713ce1f12f6a216752a2b76e8f"
},
{
"dimension": "exact_span",
"from": "citation:V2-TERRA-01:s5_urlhealth",
"status": "unresolved",
"to": "span:sha256:6af594d0155420f020802bee4a4dae0cd19f46d2c69dc107c69e75e298299169"
},
{
"dimension": "exact_span",
"from": "citation:V2-TERRA-01:s5_urlhealth",
"status": "unresolved",
"to": "span:sha256:baa7644423c27da0263b753522f1c14915dfdb8f9505577f8ff755a0e668c023"
},
{
"dimension": "requested_url",
"from": "citation:V2-TERRA-01:s5_urlhealth",
"status": "retrieved",
"to": "https://arxiv.org/pdf/2604.03173"
},
{
"dimension": "source_work",
"from": "citation:V2-TERRA-01:s5_urlhealth",
"status": "normalized",
"to": "work:url-health"
},
{
"dimension": "edition",
"from": "citation:V2-TERRA-01:s5_urlhealth",
"status": "examined",
"to": "edition:url-health-arxiv-v1"
},
{
"dimension": "upstream_citation",
"from": "claim:V2-TERRA-01:r1_deepresearchbench_fact",
"status": "declared-but-citation-unresolved",
"to": "citation:V2-TERRA-01:s1_deepresearchbench"
},
{
"dimension": "upstream_citation",
"from": "claim:V2-TERRA-01:r2_reportbench_cited_statement_match",
"status": "declared-but-citation-unresolved",
"to": "citation:V2-TERRA-01:s2_reportbench"
},
{
"dimension": "upstream_citation",
"from": "claim:V2-TERRA-01:r3_researcherbench_faithfulness_and_groundedness",
"status": "declared-but-citation-unresolved",
"to": "citation:V2-TERRA-01:s3_researcherbench"
},
{
"dimension": "upstream_citation",
"from": "claim:V2-TERRA-01:r4_cited_not_verified_source_attribution",
"status": "declared-but-citation-unresolved",
"to": "citation:V2-TERRA-01:s4_cited_not_verified"
},
{
"dimension": "upstream_citation",
"from": "claim:V2-TERRA-01:r5_urlhealth_resolution",
"status": "declared-but-citation-unresolved",
"to": "citation:V2-TERRA-01:s5_urlhealth"
},
{
"dimension": "model_profile",
"from": "report:V2-TERRA-02",
"status": "requested-not-independent",
"to": "model-profile:gpt-5.6-terra"
},
{
"dimension": "prompt",
"from": "report:V2-TERRA-02",
"status": "shared-exact-bytes",
"to": "prompt:sha256:d321a9cec7b5fe419157c0623e18ff0020cb0080fbe6dd9ed4fd20a0b896f670"
},
{
"dimension": "retrieval_infrastructure",
"from": "report:V2-TERRA-02",
"status": "unknown-or-shared",
"to": "retrieval:codex-public-web-implementation-unknown"
},
{
"dimension": "exact_span",
"from": "citation:V2-TERRA-02:s1_deepresearchbench",
"status": "exact-normalized-match",
"to": "span:sha256:eaf37bef7b8dc35f3e00b85d3e58ad872dc0ea604a062c8cdd8929184587d5d9"
},
{
"dimension": "exact_span",
"from": "citation:V2-TERRA-02:s1_deepresearchbench",
"status": "unresolved",
"to": "span:sha256:b74af48901531355cd4351211eac41aef691ea4eb29488347782fca1ec2d1eb8"
},
{
"dimension": "exact_span",
"from": "citation:V2-TERRA-02:s1_deepresearchbench",
"status": "exact-normalized-match",
"to": "span:sha256:e4a9f30effc031388ed5c9993dc6f071903bec5e3f1998db4c08ab00b284b27f"
},
{
"dimension": "exact_span",
"from": "citation:V2-TERRA-02:s1_deepresearchbench",
"status": "exact-normalized-match",
"to": "span:sha256:f98f1274859ba98cc538e3be6d11d345a5382cfcb9516aa1cf3f3278ea646b92"
},
{
"dimension": "requested_url",
"from": "citation:V2-TERRA-02:s1_deepresearchbench",
"status": "retrieved",
"to": "https://deepresearch-bench.github.io/static/papers/deepresearch-bench.pdf"
},
{
"dimension": "source_work",
"from": "citation:V2-TERRA-02:s1_deepresearchbench",
"status": "normalized",
"to": "work:deepresearch-bench-paper"
},
{
"dimension": "edition",
"from": "citation:V2-TERRA-02:s1_deepresearchbench",
"status": "examined",
"to": "edition:drbench-arxiv-v1"
},
{
"dimension": "exact_span",
"from": "citation:V2-TERRA-02:s2_liveresearchbench",
"status": "ordered-fragment-match",
"to": "span:sha256:d880991ad784390029d49f84949d083c4789d67d45893d87cfb6b5a38703453f"
},
{
"dimension": "exact_span",
"from": "citation:V2-TERRA-02:s2_liveresearchbench",
"status": "unresolved",
"to": "span:sha256:256817dca0b5aac5741ec51862ec8d0f9b4b93ea78186a4041052c7b2d1c4382"
},
{
"dimension": "exact_span",
"from": "citation:V2-TERRA-02:s2_liveresearchbench",
"status": "exact-normalized-match",
"to": "span:sha256:e6f447129d9a662fb202d4ac22e16822bacfdbcbcdb7afe3b589e21802890b8a"
},
{
"dimension": "requested_url",
"from": "citation:V2-TERRA-02:s2_liveresearchbench",
"status": "retrieved",
"to": "https://arxiv.org/pdf/2510.14240v5"
},
{
"dimension": "source_work",
"from": "citation:V2-TERRA-02:s2_liveresearchbench",
"status": "normalized",
"to": "work:liveresearchbench"
},
{
"dimension": "edition",
"from": "citation:V2-TERRA-02:s2_liveresearchbench",
"status": "examined",
"to": "edition:liveresearchbench-arxiv-v5"
},
{
"dimension": "exact_span",
"from": "citation:V2-TERRA-02:s3_deeptrace",
"status": "unresolved",
"to": "span:sha256:4d92f4721bb3d81403154a23d7fde5b342251d50a2d1034ec6d4542a35e9fc8b"
},
{
"dimension": "exact_span",
"from": "citation:V2-TERRA-02:s3_deeptrace",
"status": "unresolved",
"to": "span:sha256:7572651679e7e8a0a1a6e394504e75749ae2649cfb55f7a80531c8b53acb31dc"
},
{
"dimension": "exact_span",
"from": "citation:V2-TERRA-02:s3_deeptrace",
"status": "exact-normalized-match",
"to": "span:sha256:9cb1700c08368feb234207045354cbf559d6ed80f2b0bae814d2b5b3c858b73c"
},
{
"dimension": "exact_span",
"from": "citation:V2-TERRA-02:s3_deeptrace",
"status": "exact-normalized-match",
"to": "span:sha256:aa1534ce2e177153ab85e512deb8d064d14ec1eae16d8032bdb5043252738d4c"
},
{
"dimension": "requested_url",
"from": "citation:V2-TERRA-02:s3_deeptrace",
"status": "retrieved",
"to": "https://proceedings.iclr.cc/paper_files/paper/2026/file/ad08767706825033b99122332293033d-Paper-Conference.pdf"
},
{
"dimension": "source_work",
"from": "citation:V2-TERRA-02:s3_deeptrace",
"status": "normalized",
"to": "work:deeptrace"
},
{
"dimension": "edition",
"from": "citation:V2-TERRA-02:s3_deeptrace",
"status": "examined",
"to": "edition:deeptrace-iclr-2026"
},
{
"dimension": "exact_span",
"from": "citation:V2-TERRA-02:s4_researcherbench",
"status": "exact-normalized-match",
"to": "span:sha256:886b50d94e3facabc487342cd2af5157cf956622ecfc368684596c5f6b61c350"
},
{
"dimension": "exact_span",
"from": "citation:V2-TERRA-02:s4_researcherbench",
"status": "unresolved",
"to": "span:sha256:f85edb1051cbc19ee38bc796ed3392abf5e44fc869d59d7e82bfe2cb81696e9b"
},
{
"dimension": "requested_url",
"from": "citation:V2-TERRA-02:s4_researcherbench",
"status": "retrieved",
"to": "https://arxiv.org/html/2507.16280"
},
{
"dimension": "source_work",
"from": "citation:V2-TERRA-02:s4_researcherbench",
"status": "normalized",
"to": "work:researcherbench"
},
{
"dimension": "edition",
"from": "citation:V2-TERRA-02:s4_researcherbench",
"status": "examined",
"to": "edition:researcherbench-arxiv-v1"
},
{
"dimension": "upstream_citation",
"from": "claim:V2-TERRA-02:r1_deepresearchbench_fact",
"status": "declared-but-citation-unresolved",
"to": "citation:V2-TERRA-02:s1_deepresearchbench"
},
{
"dimension": "upstream_citation",
"from": "claim:V2-TERRA-02:r2_liveresearchbench_wide_info",
"status": "declared-but-citation-unresolved",
"to": "citation:V2-TERRA-02:s2_liveresearchbench"
},
{
"dimension": "upstream_citation",
"from": "claim:V2-TERRA-02:r3_liveresearchbench_market_analysis",
"status": "declared-but-citation-unresolved",
"to": "citation:V2-TERRA-02:s2_liveresearchbench"
},
{
"dimension": "upstream_citation",
"from": "claim:V2-TERRA-02:r4_deeptrace",
"status": "declared-but-citation-unresolved",
"to": "citation:V2-TERRA-02:s3_deeptrace"
},
{
"dimension": "upstream_citation",
"from": "claim:V2-TERRA-02:r5_researcherbench",
"status": "declared-but-citation-unresolved",
"to": "citation:V2-TERRA-02:s4_researcherbench"
},
{
"dimension": "model_profile",
"from": "report:V2-TERRA-03",
"status": "requested-not-independent",
"to": "model-profile:gpt-5.6-terra"
},
{
"dimension": "prompt",
"from": "report:V2-TERRA-03",
"status": "shared-exact-bytes",
"to": "prompt:sha256:d321a9cec7b5fe419157c0623e18ff0020cb0080fbe6dd9ed4fd20a0b896f670"
},
{
"dimension": "retrieval_infrastructure",
"from": "report:V2-TERRA-03",
"status": "unknown-or-shared",
"to": "retrieval:codex-public-web-implementation-unknown"
},
{
"dimension": "exact_span",
"from": "citation:V2-TERRA-03:S1",
"status": "exact-normalized-match",
"to": "span:sha256:cd21c450c0d33df20a2b1a539d5725347db1b5bab7d1e9866c24790d845e5873"
},
{
"dimension": "exact_span",
"from": "citation:V2-TERRA-03:S1",
"status": "unresolved",
"to": "span:sha256:35357cd8e4ff1531a84b4a5c8923db97c7b44fecd19d7666f326bbf7a1d5ec45"
},
{
"dimension": "exact_span",
"from": "citation:V2-TERRA-03:S1",
"status": "unresolved",
"to": "span:sha256:d004f718815ff769195d087f1159e50cf5d00261b8ef85229625a65740d55b3d"
},
{
"dimension": "exact_span",
"from": "citation:V2-TERRA-03:S1",
"status": "unresolved",
"to": "span:sha256:ec8d8cf5ad9b690ecd983e4d5c673d9624200351579e72868291abc15d0f2949"
},
{
"dimension": "exact_span",
"from": "citation:V2-TERRA-03:S1",
"status": "unresolved",
"to": "span:sha256:fe6b18fa48b171b5bd7c74fda45f38ed1b087ee5035f6d35dcf8c701b3033ea8"
},
{
"dimension": "exact_span",
"from": "citation:V2-TERRA-03:S1",
"status": "ordered-fragment-match",
"to": "span:sha256:5cd9ba4ca72da10cf51c0beddc6ea2765c74721bdd526a67156f44754219f440"
},
{
"dimension": "requested_url",
"from": "citation:V2-TERRA-03:S1",
"status": "retrieved",
"to": "https://arxiv.org/html/2605.06635"
},
{
"dimension": "source_work",
"from": "citation:V2-TERRA-03:S1",
"status": "normalized",
"to": "work:cited-not-verified"
},
{
"dimension": "edition",
"from": "citation:V2-TERRA-03:S1",
"status": "examined",
"to": "edition:cnv-arxiv-v1"
},
{
"dimension": "exact_span",
"from": "citation:V2-TERRA-03:S2",
"status": "exact-normalized-match",
"to": "span:sha256:015bb68c904bf225e84c82b1acffb3cc38a4a087d7b0f79d3b0102b685b2265e"
},
{
"dimension": "exact_span",
"from": "citation:V2-TERRA-03:S2",
"status": "ordered-fragment-match",
"to": "span:sha256:0d4f357f2a5565bdaea01a1e537650958d71216868d1a2b0bc14c47a45b6d0ef"
},
{
"dimension": "exact_span",
"from": "citation:V2-TERRA-03:S2",
"status": "unresolved",
"to": "span:sha256:3ec266d89a7c774ebcee3d2b8b758ac56176c382481cd0d2f0bd893aaa142281"
},
{
"dimension": "exact_span",
"from": "citation:V2-TERRA-03:S2",
"status": "unresolved",
"to": "span:sha256:de53fab3c5b2e262ba56cdd5a27eb002c2b4641755ccc38c734a5bd506a670ef"
},
{
"dimension": "requested_url",
"from": "citation:V2-TERRA-03:S2",
"status": "retrieved",
"to": "https://arxiv.org/html/2506.11763"
},
{
"dimension": "source_work",
"from": "citation:V2-TERRA-03:S2",
"status": "normalized",
"to": "work:deepresearch-bench-paper"
},
{
"dimension": "edition",
"from": "citation:V2-TERRA-03:S2",
"status": "examined",
"to": "edition:drbench-arxiv-v1"
},
{
"dimension": "exact_span",
"from": "citation:V2-TERRA-03:S3",
"status": "unresolved",
"to": "span:sha256:42a4184b2e116330dc263a3d8e9b77f4b88c9da1c325fc67c4375df0f818a514"
},
{
"dimension": "exact_span",
"from": "citation:V2-TERRA-03:S3",
"status": "exact-normalized-match",
"to": "span:sha256:ba6b8a8b7d9121ebb05d594176ceed3d1e9dbc637876783abfd0141ebbd0a822"
},
{
"dimension": "exact_span",
"from": "citation:V2-TERRA-03:S3",
"status": "unresolved",
"to": "span:sha256:7eefee71f11e6f5ec8c2e70272fd9189464650e68b904d9fa9c6d059bedc4195"
},
{
"dimension": "exact_span",
"from": "citation:V2-TERRA-03:S3",
"status": "exact-normalized-match",
"to": "span:sha256:1db12b68cfeb4c5f5a96fa9cd2e2bd46e8dfb4f096f73154481d29e22a4d5414"
},
{
"dimension": "requested_url",
"from": "citation:V2-TERRA-03:S3",
"status": "retrieved",
"to": "https://arxiv.org/html/2508.15804"
},
{
"dimension": "source_work",
"from": "citation:V2-TERRA-03:S3",
"status": "normalized",
"to": "work:reportbench"
},
{
"dimension": "edition",
"from": "citation:V2-TERRA-03:S3",
"status": "examined",
"to": "edition:reportbench-arxiv-v1"
},
{
"dimension": "exact_span",
"from": "citation:V2-TERRA-03:S4",
"status": "exact-normalized-match",
"to": "span:sha256:b38a2af5982b4c85bc205d2f533a23ed3a8f40a49b651a5f00ec6932dcc67f7d"
},
{
"dimension": "exact_span",
"from": "citation:V2-TERRA-03:S4",
"status": "exact-normalized-match",
"to": "span:sha256:b6a6ba48a2c4e352c8565b671a15a90400739f2aeefda8a65a31ad8a1ff9a4e4"
},
{
"dimension": "exact_span",
"from": "citation:V2-TERRA-03:S4",
"status": "unresolved",
"to": "span:sha256:f8ff0887370fedb6d6676d1eaf59f86461fe5083d20dfa5b28f8ba94175ed031"
},
{
"dimension": "requested_url",
"from": "citation:V2-TERRA-03:S4",
"status": "retrieved",
"to": "https://pmc.ncbi.nlm.nih.gov/articles/PMC13109748/"
},
{
"dimension": "source_work",
"from": "citation:V2-TERRA-03:S4",
"status": "normalized",
"to": "work:keplinger-dermatology-audit"
},
{
"dimension": "edition",
"from": "citation:V2-TERRA-03:S4",
"status": "examined",
"to": "edition:keplinger-vor-2025"
},
{
"dimension": "exact_span",
"from": "citation:V2-TERRA-03:S5",
"status": "ordered-fragment-match",
"to": "span:sha256:ef637b38f36195bea09205d532e8744daf3e257f93d97144f061024092bf67fe"
},
{
"dimension": "exact_span",
"from": "citation:V2-TERRA-03:S5",
"status": "exact-normalized-match",
"to": "span:sha256:92eb7cf19745e24190dda84c77d590941134d09f06805c0fb605207fd942ae11"
},
{
"dimension": "requested_url",
"from": "citation:V2-TERRA-03:S5",
"status": "retrieved",
"to": "https://github.com/Ayanami0730/deep_research_bench"
},
{
"dimension": "source_work",
"from": "citation:V2-TERRA-03:S5",
"status": "normalized",
"to": "work:deepresearch-bench-repository"
},
{
"dimension": "edition",
"from": "citation:V2-TERRA-03:S5",
"status": "examined",
"to": "edition:drbench-repo-main-469cce5"
},
{
"dimension": "exact_span",
"from": "citation:V2-TERRA-03:S6",
"status": "unresolved",
"to": "span:sha256:fdc35d731d751e7e0691158d8fed4a2934851897f9fef016b33376cde973827a"
},
{
"dimension": "requested_url",
"from": "citation:V2-TERRA-03:S6",
"status": "inaccessible",
"to": "https://pubmed.ncbi.nlm.nih.gov/40904191/"
},
{
"dimension": "source_work",
"from": "citation:V2-TERRA-03:S6",
"status": "normalized",
"to": "work:keplinger-dermatology-audit"
},
{
"dimension": "edition",
"from": "citation:V2-TERRA-03:S6",
"status": "examined",
"to": "edition:keplinger-vor-2025"
},
{
"dimension": "upstream_citation",
"from": "claim:V2-TERRA-03:R1",
"status": "declared-but-citation-unresolved",
"to": "citation:V2-TERRA-03:S1"
},
{
"dimension": "upstream_citation",
"from": "claim:V2-TERRA-03:R2",
"status": "declared-but-citation-unresolved",
"to": "citation:V2-TERRA-03:S1"
},
{
"dimension": "upstream_citation",
"from": "claim:V2-TERRA-03:R3",
"status": "declared-but-citation-unresolved",
"to": "citation:V2-TERRA-03:S2"
},
{
"dimension": "upstream_citation",
"from": "claim:V2-TERRA-03:R3",
"status": "declared-and-resolved",
"to": "citation:V2-TERRA-03:S5"
},
{
"dimension": "upstream_citation",
"from": "claim:V2-TERRA-03:R4",
"status": "declared-but-citation-unresolved",
"to": "citation:V2-TERRA-03:S3"
},
{
"dimension": "upstream_citation",
"from": "claim:V2-TERRA-03:R5",
"status": "declared-but-citation-unresolved",
"to": "citation:V2-TERRA-03:S4"
},
{
"dimension": "upstream_citation",
"from": "claim:V2-TERRA-03:R5",
"status": "declared-but-citation-unresolved",
"to": "citation:V2-TERRA-03:S6"
},
{
"dimension": "model_profile",
"from": "report:V2-TERRA-04",
"status": "requested-not-independent",
"to": "model-profile:gpt-5.6-terra"
},
{
"dimension": "prompt",
"from": "report:V2-TERRA-04",
"status": "shared-exact-bytes",
"to": "prompt:sha256:d321a9cec7b5fe419157c0623e18ff0020cb0080fbe6dd9ed4fd20a0b896f670"
},
{
"dimension": "retrieval_infrastructure",
"from": "report:V2-TERRA-04",
"status": "unknown-or-shared",
"to": "retrieval:codex-public-web-implementation-unknown"
},
{
"dimension": "exact_span",
"from": "citation:V2-TERRA-04:S1_deepresearchbench",
"status": "exact-normalized-match",
"to": "span:sha256:9fe42cb48702d8d96ebbbc0309692214ece6ec495c447c06e2be62f32ff00894"
},
{
"dimension": "exact_span",
"from": "citation:V2-TERRA-04:S1_deepresearchbench",
"status": "unresolved",
"to": "span:sha256:539d8123b83aa1250f16cb7a8a4a66ece5d42099c0ae7c7d1fd877f1fbafb3bd"
},
{
"dimension": "exact_span",
"from": "citation:V2-TERRA-04:S1_deepresearchbench",
"status": "unresolved",
"to": "span:sha256:a13bd114b7a394950897d477dccb999910794b70914bbc3a45fa6c4667181992"
},
{
"dimension": "exact_span",
"from": "citation:V2-TERRA-04:S1_deepresearchbench",
"status": "exact-normalized-match",
"to": "span:sha256:76e9bd59b31c7ebd2ed7a2186e59753a79d3c730e0d7b4fbf0146fb39f4e9c5f"
},
{
"dimension": "requested_url",
"from": "citation:V2-TERRA-04:S1_deepresearchbench",
"status": "retrieved",
"to": "https://arxiv.org/html/2506.11763"
},
{
"dimension": "source_work",
"from": "citation:V2-TERRA-04:S1_deepresearchbench",
"status": "normalized",
"to": "work:deepresearch-bench-paper"
},
{
"dimension": "edition",
"from": "citation:V2-TERRA-04:S1_deepresearchbench",
"status": "examined",
"to": "edition:drbench-arxiv-v1"
},
{
"dimension": "exact_span",
"from": "citation:V2-TERRA-04:S2_researcherbench",
"status": "exact-normalized-match",
"to": "span:sha256:f96ad5daa81153a3689f52f5777fedfd70e8a87d44e7a1c9e2c22b5789a3707d"
},
{
"dimension": "exact_span",
"from": "citation:V2-TERRA-04:S2_researcherbench",
"status": "unresolved",
"to": "span:sha256:5dbd4e3e48480f5b3c54ce606082c35bfd55bf40cd0be847cc03a8fd4d9993f6"
},
{
"dimension": "exact_span",
"from": "citation:V2-TERRA-04:S2_researcherbench",
"status": "unresolved",
"to": "span:sha256:0d70187dbf9e3044ae9efd387faf34b29b9ab556aae6b0ea4c06e05964bfa2c7"
},
{
"dimension": "exact_span",
"from": "citation:V2-TERRA-04:S2_researcherbench",
"status": "exact-normalized-match",
"to": "span:sha256:ea9cb2e1b0078c0857b789d61d0a0d26801cf6689186470588a87b98b4639796"
},
{
"dimension": "requested_url",
"from": "citation:V2-TERRA-04:S2_researcherbench",
"status": "retrieved",
"to": "https://arxiv.org/html/2507.16280"
},
{
"dimension": "source_work",
"from": "citation:V2-TERRA-04:S2_researcherbench",
"status": "normalized",
"to": "work:researcherbench"
},
{
"dimension": "edition",
"from": "citation:V2-TERRA-04:S2_researcherbench",
"status": "examined",
"to": "edition:researcherbench-arxiv-v1"
},
{
"dimension": "exact_span",
"from": "citation:V2-TERRA-04:S3_reportbench",
"status": "exact-normalized-match",
"to": "span:sha256:9c7a4c186cc85f185aa293a7c5a46c08f9dbbc1a9e63c449dc737447e42f8ee4"
},
{
"dimension": "exact_span",
"from": "citation:V2-TERRA-04:S3_reportbench",
"status": "unresolved",
"to": "span:sha256:9a2cf22761f2ebb97eb31dfa02a32775248c3a7396b1e410565c313d30585cf3"
},
{
"dimension": "exact_span",
"from": "citation:V2-TERRA-04:S3_reportbench",
"status": "exact-normalized-match",
"to": "span:sha256:ddf4320a6d302c13d28b6225eb7e79ac15583349e1b449d4bb967eb89a257a3d"
},
{
"dimension": "exact_span",
"from": "citation:V2-TERRA-04:S3_reportbench",
"status": "exact-normalized-match",
"to": "span:sha256:084f11e57cf779850ea87aaf30ba1277438ff9efcb533fd2ebb09dba3665be8d"
},
{
"dimension": "requested_url",
"from": "citation:V2-TERRA-04:S3_reportbench",
"status": "retrieved",
"to": "https://arxiv.org/html/2508.15804"
},
{
"dimension": "source_work",
"from": "citation:V2-TERRA-04:S3_reportbench",
"status": "normalized",
"to": "work:reportbench"
},
{
"dimension": "edition",
"from": "citation:V2-TERRA-04:S3_reportbench",
"status": "examined",
"to": "edition:reportbench-arxiv-v1"
},
{
"dimension": "exact_span",
"from": "citation:V2-TERRA-04:S4_urlhealth",
"status": "unresolved",
"to": "span:sha256:131799471acf2f83cc3a17322a1c5cda6b5b46020bd894be54a1708db62eb14b"
},
{
"dimension": "exact_span",
"from": "citation:V2-TERRA-04:S4_urlhealth",
"status": "unresolved",
"to": "span:sha256:dc108a5e00fb6ce5306fcde8fd0c36c85352dac361662fab3ad1eea89d66c048"
},
{
"dimension": "exact_span",
"from": "citation:V2-TERRA-04:S4_urlhealth",
"status": "exact-normalized-match",
"to": "span:sha256:bb62928b5e0cb4e373b2f6bfeb579d8564d05da5588f8b150a89ea4b9b2ed0db"
},
{
"dimension": "exact_span",
"from": "citation:V2-TERRA-04:S4_urlhealth",
"status": "exact-normalized-match",
"to": "span:sha256:59d0da64f738e1d8745f2bff3663eb89e273f93fe501afe37e2494a8e0ff31a1"
},
{
"dimension": "requested_url",
"from": "citation:V2-TERRA-04:S4_urlhealth",
"status": "retrieved",
"to": "https://arxiv.org/html/2604.03173"
},
{
"dimension": "source_work",
"from": "citation:V2-TERRA-04:S4_urlhealth",
"status": "normalized",
"to": "work:url-health"
},
{
"dimension": "edition",
"from": "citation:V2-TERRA-04:S4_urlhealth",
"status": "examined",
"to": "edition:url-health-arxiv-v1"
},
{
"dimension": "exact_span",
"from": "citation:V2-TERRA-04:S5_cited_not_verified",
"status": "exact-normalized-match",
"to": "span:sha256:732c03e69c6e142a492275cee5d1e2d474c75422c9b84aaaf273100e937338ed"
},
{
"dimension": "exact_span",
"from": "citation:V2-TERRA-04:S5_cited_not_verified",
"status": "unresolved",
"to": "span:sha256:cebc739ec2bf421e0853467bfeccb3f42bda1cf05472076a1c1bc6d9fb1edfb9"
},
{
"dimension": "exact_span",
"from": "citation:V2-TERRA-04:S5_cited_not_verified",
"status": "unresolved",
"to": "span:sha256:caaa3e2fbb4cef64558f2c8cc30388281e1505c9353fb775f960a5f18c816111"
},
{
"dimension": "exact_span",
"from": "citation:V2-TERRA-04:S5_cited_not_verified",
"status": "exact-normalized-match",
"to": "span:sha256:d638c8b21f0e4a0f840880b6ffe0217ce1623758cc09fdccc32cacf2db3dd958"
},
{
"dimension": "exact_span",
"from": "citation:V2-TERRA-04:S5_cited_not_verified",
"status": "ordered-fragment-match",
"to": "span:sha256:c7dbd1d9e09980703ecbee8161594e4b426c1af63117b28662a7e98686107621"
},
{
"dimension": "requested_url",
"from": "citation:V2-TERRA-04:S5_cited_not_verified",
"status": "retrieved",
"to": "https://arxiv.org/html/2605.06635"
},
{
"dimension": "source_work",
"from": "citation:V2-TERRA-04:S5_cited_not_verified",
"status": "normalized",
"to": "work:cited-not-verified"
},
{
"dimension": "edition",
"from": "citation:V2-TERRA-04:S5_cited_not_verified",
"status": "examined",
"to": "edition:cnv-arxiv-v1"
},
{
"dimension": "upstream_citation",
"from": "claim:V2-TERRA-04:R1_deepresearchbench_support",
"status": "declared-but-citation-unresolved",
"to": "citation:V2-TERRA-04:S1_deepresearchbench"
},
{
"dimension": "upstream_citation",
"from": "claim:V2-TERRA-04:R2_researcherbench_support",
"status": "declared-but-citation-unresolved",
"to": "citation:V2-TERRA-04:S2_researcherbench"
},
{
"dimension": "upstream_citation",
"from": "claim:V2-TERRA-04:R3_reportbench_support",
"status": "declared-but-citation-unresolved",
"to": "citation:V2-TERRA-04:S3_reportbench"
},
{
"dimension": "upstream_citation",
"from": "claim:V2-TERRA-04:R4_urlhealth_resolution",
"status": "declared-but-citation-unresolved",
"to": "citation:V2-TERRA-04:S4_urlhealth"
},
{
"dimension": "upstream_citation",
"from": "claim:V2-TERRA-04:R5_cited_not_verified_support_and_resolution",
"status": "declared-but-citation-unresolved",
"to": "citation:V2-TERRA-04:S5_cited_not_verified"
},
{
"dimension": "upstream_citation",
"from": "claim:V2-TERRA-04:R6_search_depth_ablation",
"status": "declared-but-citation-unresolved",
"to": "citation:V2-TERRA-04:S5_cited_not_verified"
}
],
"derived_counts": {
"candidate_warrant_roots": 7,
"captured_reports": 8,
"citation_occurrences": 48,
"cited_urls": 30,
"exact_span_roots": 72,
"examined_edition_roots": 14,
"inaccessible_citations": 3,
"independently_confirmed_warrant_roots": 0,
"invalid_citations": 0,
"non_resolving_citations": 0,
"raw_claim_occurrences": 52,
"raw_span_occurrences": 127,
"resolving_url_roots": 27,
"source_work_roots": 11,
"unresolved_citations": 34,
"unsupported_or_force_raised_claims": 20
},
"disposition_counts": {
"em0029_review_supplement_spans": 3,
"independently_rejected_claim_occurrences": 9,
"pending_warrant_roots": 4
},
"examined_edition_ids": [
"edition:citation-verifier-arxiv-v1",
"edition:cnv-arxiv-v1",
"edition:deeptrace-arxiv-v1",
"edition:deeptrace-iclr-2026",
"edition:drbench-arxiv-v1",
"edition:drbench-iclr-2026",
"edition:drbench-repo-main-469cce5",
"edition:keplinger-supplement-v1",
"edition:keplinger-supplement-v2",
"edition:keplinger-vor-2025",
"edition:liveresearchbench-arxiv-v5",
"edition:reportbench-arxiv-v1",
"edition:researcherbench-arxiv-v1",
"edition:url-health-arxiv-v1"
],
"format": "epistemedia-case-002-audit-projection-v0.1",
"inaccessible_citations": [
{
"citation_occurrence_id": "V2-SOL-02:S4",
"correction_ids": [],
"edition_id": "edition:drbench-iclr-2026",
"license": "unknown",
"license_treatment": "metadata, digest, and quote-minimal attributed spans only",
"raw_source_id": "S4",
"raw_title": "DeepResearch Bench: A Comprehensive Benchmark for Deep Research Agents",
"readback": {
"captured_bytes": 12692,
"captured_sha256": "a2d811e0d24af89d9bd4b19cc9e8fbc474c67ef3827648550a6560eca2a5e332",
"edition_id": "edition:drbench-iclr-2026",
"http_status": 403,
"media_type": "text/html",
"redirect_chain": [],
"requested_url": "https://openreview.net/pdf?id=hQ0K2Hhq7H",
"resolved_url": "https://openreview.net/pdf?id=hQ0K2Hhq7H",
"retrieval_status": "inaccessible"
},
"requested_url": "https://openreview.net/pdf?id=hQ0K2Hhq7H",
"resolution_status": "unresolved",
"run_id": "V2-SOL-02",
"source_work_id": "work:deepresearch-bench-paper",
"span_occurrence_ids": [
"V2-SOL-02:S4:S4_SPAN_1",
"V2-SOL-02:S4:S4_SPAN_2"
]
},
{
"citation_occurrence_id": "V2-SOL-03:s1",
"correction_ids": [],
"edition_id": "edition:keplinger-vor-2025",
"license": "CC BY-NC 4.0",
"license_treatment": "quote-minimal attributed spans; no full-text redistribution",
"raw_source_id": "s1",
"raw_title": "Assessment of Deep Research for dermatology literature reviews: Deep concern over the hype",
"readback": {
"captured_bytes": 5498,
"captured_sha256": "d74c1891ae09a2b89121f2a60ac837c89761a8086759e45db214ca68c1905cea",
"edition_id": "edition:keplinger-vor-2025",
"http_status": 403,
"media_type": "text/html; charset=UTF-8",
"redirect_chain": [],
"requested_url": "https://onlinelibrary.wiley.com/doi/10.1111/jdv.70035",
"resolved_url": "https://onlinelibrary.wiley.com/doi/10.1111/jdv.70035",
"retrieval_status": "inaccessible"
},
"requested_url": "https://onlinelibrary.wiley.com/doi/10.1111/jdv.70035",
"resolution_status": "unresolved",
"run_id": "V2-SOL-03",
"source_work_id": "work:keplinger-dermatology-audit",
"span_occurrence_ids": [
"V2-SOL-03:s1:s1_span1"
]
},
{
"citation_occurrence_id": "V2-TERRA-03:S6",
"correction_ids": [],
"edition_id": "edition:keplinger-vor-2025",
"license": "CC BY-NC 4.0",
"license_treatment": "quote-minimal attributed spans; no full-text redistribution",
"raw_source_id": "S6",
"raw_title": "PubMed record and Figure 1 caption for Assessment of Deep Research for dermatology literature reviews: Deep concern over the hype",
"readback": {
"captured_bytes": 5565,
"captured_sha256": "a46109544fe4ff4504fab5e97abea3cb7172367aba54a308937386394f0ff046",
"edition_id": "edition:keplinger-vor-2025",
"http_status": 203,
"media_type": "text/html; charset=utf-8",
"redirect_chain": [],
"requested_url": "https://pubmed.ncbi.nlm.nih.gov/40904191/",
"resolved_url": "https://pubmed.ncbi.nlm.nih.gov/40904191/",
"retrieval_status": "inaccessible"
},
"requested_url": "https://pubmed.ncbi.nlm.nih.gov/40904191/",
"resolution_status": "unresolved",
"run_id": "V2-TERRA-03",
"source_work_id": "work:keplinger-dermatology-audit",
"span_occurrence_ids": [
"V2-TERRA-03:S6:S6-A"
]
}
],
"independently_confirmed_warrant_ids": [],
"limitations": [
"The eight reports share exact prompt bytes, Codex runtime lineage, and unknown or shared retrieval infrastructure; they are observations, not independent evidence roots.",
"Candidate warrant roots are author-side normalization results and remain zero independently confirmed roots until a fresh-clone reviewer repeats source and count checks.",
"HTTP readback, text matching, and semantic warrant are separate gates.",
"No universal claim about all agents, current products, or all web sources is authorized."
],
"matched_span_root_ids": [
"span:sha256:015bb68c904bf225e84c82b1acffb3cc38a4a087d7b0f79d3b0102b685b2265e",
"span:sha256:084f11e57cf779850ea87aaf30ba1277438ff9efcb533fd2ebb09dba3665be8d",
"span:sha256:092745c7db6407d4d52927642741cde229fad05a5347bb1867a5154421060e70",
"span:sha256:0abbab3d6269d58eb76d8e90eff1e4c27512cf46d9b8e5ff0a9f537b18f232fa",
"span:sha256:0d4f357f2a5565bdaea01a1e537650958d71216868d1a2b0bc14c47a45b6d0ef",
"span:sha256:13cb0ddb087b82988b7445a9cfb02562f9c044ef4f244e8b176ff475859eaea3",
"span:sha256:196c54d6212fbc0c054dbf8e8467f2388c1f788bb93f4b1a4b6ae085eed17685",
"span:sha256:1db12b68cfeb4c5f5a96fa9cd2e2bd46e8dfb4f096f73154481d29e22a4d5414",
"span:sha256:279db725380c133044944dccb379b75a396242219aa30437b26ddc40b09a3b5c",
"span:sha256:2f311bfb451c6208101cb7583af37c338914fcdb10f5c75b131e23d5bb08e66a",
"span:sha256:3015a109e7898cb720e426e9a57c526277b648c75f1fb4b13631b00524eac943",
"span:sha256:35927851edd2e7e5e0f60d98498940f88304ba99bf1f85a08535663943c2f40d",
"span:sha256:39a9931f05ea356fe904ed15768fa08c66440b1fcb20f98e267a315e6d562af9",
"span:sha256:417182b4c15ac550469e435a02aab1e91b395f72f186312de9d34d57fd1c6c98",
"span:sha256:46c392bc942cd88d525d74298e6bc6ebd9aeaec533ddfb7b7b761a9f5b454b05",
"span:sha256:55de6a6ca9414a252592e9f77544a9259533dd713ce1f12f6a216752a2b76e8f",
"span:sha256:59d0da64f738e1d8745f2bff3663eb89e273f93fe501afe37e2494a8e0ff31a1",
"span:sha256:5cd9ba4ca72da10cf51c0beddc6ea2765c74721bdd526a67156f44754219f440",
"span:sha256:6885cf2524462022475b5829da081b2d35fcfe7463ca6ad0979e7c4fb1f0b54d",
"span:sha256:6c1d327e58ef001533ea1c11010e974436bbeafbde138efcc7b516281f6d25a5",
"span:sha256:6e961498ed0f81a950f775956587f741e0eb9f11640fd3cb40870585558d869d",
"span:sha256:6f79c3c824ccf889d73e234335661e5fca68d90f8c78ee97aef488e4fdb32a49",
"span:sha256:732c03e69c6e142a492275cee5d1e2d474c75422c9b84aaaf273100e937338ed",
"span:sha256:74f6acafbb4c54d556a572460bf1b45bacd1e293f4aa16cc2d499174096b9c20",
"span:sha256:76e9bd59b31c7ebd2ed7a2186e59753a79d3c730e0d7b4fbf0146fb39f4e9c5f",
"span:sha256:794bdf7e0595799afbb0d55f3dc522077aed9615c59f1e3575707148cf444bcf",
"span:sha256:7a6046a756d6ec0315d330cf0817d062ea48498bdc9c90b3489d64b2bd12a3a4",
"span:sha256:7d7331942f9d0520e18e96206e7718e5187788be4d48cef995fac79b235db757",
"span:sha256:886b50d94e3facabc487342cd2af5157cf956622ecfc368684596c5f6b61c350",
"span:sha256:8c8d9a6dc003bf0298542b85df3f6f08db631e68e89f606f4e53e36e86dde167",
"span:sha256:8cce3d9ee1e1674ba1b93321dd9b99a55d834ca980451f8ab61bcaf5bdda8e4d",
"span:sha256:912c931063109f84ea88aea34891dd5e0f9147d2176af258059246431e9abfb1",
"span:sha256:92eb7cf19745e24190dda84c77d590941134d09f06805c0fb605207fd942ae11",
"span:sha256:94ad583514aa7e04b8d5f7b8f9cdda6fdffe8ca6712f033da3ce4f18fe7c03f1",
"span:sha256:9c7a4c186cc85f185aa293a7c5a46c08f9dbbc1a9e63c449dc737447e42f8ee4",
"span:sha256:9cb1700c08368feb234207045354cbf559d6ed80f2b0bae814d2b5b3c858b73c",
"span:sha256:9dd6eb46529e77a62748093528e52685fbd92d46a756aef750d22aa25981db23",
"span:sha256:9fe42cb48702d8d96ebbbc0309692214ece6ec495c447c06e2be62f32ff00894",
"span:sha256:a08468cb62f2feb21d9cbb95175ce00f54592bd83573cc9166e2375947f62fee",
"span:sha256:a15ab02e9fcb23df02fb03d28967d4220ddf65ad380d223049f22b6ab68ed5eb",
"span:sha256:a18e778fac3d49e81167f05e09fbc361e91af8c0c01b0fec52991f0a1e3d0b91",
"span:sha256:aa1534ce2e177153ab85e512deb8d064d14ec1eae16d8032bdb5043252738d4c",
"span:sha256:b0a39c963cdfdc5f9852cfc9f7e33e05d31232bd1d8998851e879c38e6c34843",
"span:sha256:b38a2af5982b4c85bc205d2f533a23ed3a8f40a49b651a5f00ec6932dcc67f7d",
"span:sha256:b68153666e4e12aee8dc887fef0a61fa5f44b5f680a0b7eb95ff3defcf8b5bf1",
"span:sha256:b6a6ba48a2c4e352c8565b671a15a90400739f2aeefda8a65a31ad8a1ff9a4e4",
"span:sha256:b83996e3b5fdf878e04d6d41d0e7a1eebee9fdd9bbccf2ae4eaf33524fcfca44",
"span:sha256:b903dc2284e1b4d80bb6ef948d79bbdbc7d7447e871281b8577284883f75ccfa",
"span:sha256:ba6b8a8b7d9121ebb05d594176ceed3d1e9dbc637876783abfd0141ebbd0a822",
"span:sha256:bb16d5f06abe8641d9ea6694e549aa15e240dd88946aa47a2d4e30ddd3351a24",
"span:sha256:bb62928b5e0cb4e373b2f6bfeb579d8564d05da5588f8b150a89ea4b9b2ed0db",
"span:sha256:c27aaca02bda06ed764be53351158fc862af9d4a556d3e82074507436f48c3f8",
"span:sha256:c3dfc98566c27561eb7267972e00e3e1ba6a399a3ff5d002dca2d1233836790c",
"span:sha256:c7dbd1d9e09980703ecbee8161594e4b426c1af63117b28662a7e98686107621",
"span:sha256:cd21c450c0d33df20a2b1a539d5725347db1b5bab7d1e9866c24790d845e5873",
"span:sha256:d638c8b21f0e4a0f840880b6ffe0217ce1623758cc09fdccc32cacf2db3dd958",
"span:sha256:d880991ad784390029d49f84949d083c4789d67d45893d87cfb6b5a38703453f",
"span:sha256:ddf4320a6d302c13d28b6225eb7e79ac15583349e1b449d4bb967eb89a257a3d",
"span:sha256:deada8bacb987a47f2a42e52fbfeb5c1beb3d250c0be79e2f9c5cdd6e16b4bc7",
"span:sha256:e223640cb0ed12c87ca1d2406f3276f30a3b8d2017dd4dc1457fe59a94471040",
"span:sha256:e2c9f6220843a04665f5d1cd142ac15f04211ace5c579fa67f77e74e84a76dfe",
"span:sha256:e4a9f30effc031388ed5c9993dc6f071903bec5e3f1998db4c08ab00b284b27f",
"span:sha256:e54ee1457854a9a5e46f7c7c2e66ac6a5a69b963f8ee3450f40052f138d5d0c1",
"span:sha256:e5bcf0241a1aeae7d9a92755b1bcf0adeeea93a8922f22f832693c61780386e7",
"span:sha256:e6f447129d9a662fb202d4ac22e16822bacfdbcbcdb7afe3b589e21802890b8a",
"span:sha256:ea9cb2e1b0078c0857b789d61d0a0d26801cf6689186470588a87b98b4639796",
"span:sha256:eaf37bef7b8dc35f3e00b85d3e58ad872dc0ea604a062c8cdd8929184587d5d9",
"span:sha256:ef637b38f36195bea09205d532e8744daf3e257f93d97144f061024092bf67fe",
"span:sha256:f05b3189dc3fafd45b5bde643a91fd0616832928215a37a2430a5a658e592029",
"span:sha256:f35c9565f2fd047f6e7272dc608bbf6a63813a717bad5354c24bdaac65e40bfe",
"span:sha256:f96ad5daa81153a3689f52f5777fedfd70e8a87d44e7a1c9e2c22b5789a3707d",
"span:sha256:f98f1274859ba98cc538e3be6d11d345a5382cfcb9516aa1cf3f3278ea646b92"
],
"pending_warrant_ids": [
"warrant:citation-verifier-calibration",
"warrant:liveresearchbench-e1-e2-e3",
"warrant:researcherbench-faithfulness-groundedness",
"warrant:url-health-correction-loop"
],
"raw_claim_occurrence_ids": [
"V2-SOL-01:r1_human_dermatology_chatgpt",
"V2-SOL-01:r2_human_dermatology_lechat",
"V2-SOL-01:r3_deepresearch_bench_fact",
"V2-SOL-01:r4_deeptrace_support",
"V2-SOL-01:r5_cited_but_not_verified_cross_model",
"V2-SOL-01:r6_cited_but_not_verified_depth_ablation",
"V2-SOL-01:r7_url_resolution_drbench",
"V2-SOL-02:R1_url_resolution",
"V2-SOL-02:R2_human_dermatology_claim_support",
"V2-SOL-02:R3_deepresearch_bench_fact",
"V2-SOL-02:R4_reportbench_match_rate",
"V2-SOL-02:R5_deeptrace_audit",
"V2-SOL-02:R6_surface_quality_vs_fact_support",
"V2-SOL-02:R7_search_depth_ablation",
"V2-SOL-03:r1_reference_metadata_human_audit",
"V2-SOL-03:r2_claim_support_human_audit",
"V2-SOL-03:r3_deepresearch_bench_fact",
"V2-SOL-03:r4_deeptrace_support",
"V2-SOL-03:r5_deep_agent_url_resolution",
"V2-SOL-03:r6_url_self_correction",
"V2-SOL-03:r7_source_attribution_framework",
"V2-SOL-03:r8_search_depth_ablation",
"V2-SOL-03:r9_verifier_calibration",
"V2-SOL-04:R1",
"V2-SOL-04:R2",
"V2-SOL-04:R3",
"V2-SOL-04:R4",
"V2-SOL-04:R5",
"V2-SOL-04:R6",
"V2-SOL-04:R7",
"V2-SOL-04:R8",
"V2-TERRA-01:r1_deepresearchbench_fact",
"V2-TERRA-01:r2_reportbench_cited_statement_match",
"V2-TERRA-01:r3_researcherbench_faithfulness_and_groundedness",
"V2-TERRA-01:r4_cited_not_verified_source_attribution",
"V2-TERRA-01:r5_urlhealth_resolution",
"V2-TERRA-02:r1_deepresearchbench_fact",
"V2-TERRA-02:r2_liveresearchbench_wide_info",
"V2-TERRA-02:r3_liveresearchbench_market_analysis",
"V2-TERRA-02:r4_deeptrace",
"V2-TERRA-02:r5_researcherbench",
"V2-TERRA-03:R1",
"V2-TERRA-03:R2",
"V2-TERRA-03:R3",
"V2-TERRA-03:R4",
"V2-TERRA-03:R5",
"V2-TERRA-04:R1_deepresearchbench_support",
"V2-TERRA-04:R2_researcherbench_support",
"V2-TERRA-04:R3_reportbench_support",
"V2-TERRA-04:R4_urlhealth_resolution",
"V2-TERRA-04:R5_cited_not_verified_support_and_resolution",
"V2-TERRA-04:R6_search_depth_ablation"
],
"raw_span_occurrence_ids": [
"V2-SOL-01:s1_keplinger:sp1a",
"V2-SOL-01:s1_keplinger:sp1b",
"V2-SOL-01:s1_keplinger:sp1c",
"V2-SOL-01:s1a_keplinger_data:sp1d",
"V2-SOL-01:s2_drbench:sp2a",
"V2-SOL-01:s2_drbench:sp2b",
"V2-SOL-01:s2_drbench:sp2c",
"V2-SOL-01:s2_drbench:sp2d",
"V2-SOL-01:s2a_drbench_repo:sp2e",
"V2-SOL-01:s3_deeptrace:sp3a",
"V2-SOL-01:s3_deeptrace:sp3b",
"V2-SOL-01:s3_deeptrace:sp3c",
"V2-SOL-01:s3_deeptrace:sp3d",
"V2-SOL-01:s4_cited_not_verified:sp4a",
"V2-SOL-01:s4_cited_not_verified:sp4b",
"V2-SOL-01:s4_cited_not_verified:sp4c",
"V2-SOL-01:s5_url_health:sp5a",
"V2-SOL-01:s5_url_health:sp5b",
"V2-SOL-01:s5_url_health:sp5c",
"V2-SOL-02:S1:S1_SPAN_1",
"V2-SOL-02:S1:S1_SPAN_2",
"V2-SOL-02:S2:S2_SPAN_1",
"V2-SOL-02:S3:S3_SPAN_1",
"V2-SOL-02:S3:S3_SPAN_2",
"V2-SOL-02:S4:S4_SPAN_1",
"V2-SOL-02:S4:S4_SPAN_2",
"V2-SOL-02:S5:S5_SPAN_1",
"V2-SOL-02:S5:S5_SPAN_2",
"V2-SOL-02:S6:S6_SPAN_1",
"V2-SOL-02:S6:S6_SPAN_2",
"V2-SOL-02:S7:S7_SPAN_1",
"V2-SOL-02:S7:S7_SPAN_2",
"V2-SOL-03:s1:s1_span1",
"V2-SOL-03:s2:s2_span1",
"V2-SOL-03:s3:s3_span1",
"V2-SOL-03:s4:s4_span1",
"V2-SOL-03:s4:s4_span2",
"V2-SOL-03:s4:s4_span3",
"V2-SOL-03:s5:s5_span1",
"V2-SOL-03:s5:s5_span2",
"V2-SOL-03:s6:s6_span1",
"V2-SOL-03:s7:s7_span1",
"V2-SOL-04:S1:S1-SP1",
"V2-SOL-04:S1:S1-SP2",
"V2-SOL-04:S1:S1-SP3",
"V2-SOL-04:S2:S2-SP1",
"V2-SOL-04:S2:S2-SP2",
"V2-SOL-04:S2:S2-SP3",
"V2-SOL-04:S2:S2-SP4",
"V2-SOL-04:S3:S3-SP1",
"V2-SOL-04:S3:S3-SP2",
"V2-SOL-04:S4:S4-SP1",
"V2-SOL-04:S4:S4-SP2",
"V2-SOL-04:S5:S5-SP1",
"V2-SOL-04:S5:S5-SP2",
"V2-SOL-04:S5:S5-SP3",
"V2-SOL-04:S6:S6-SP1",
"V2-SOL-04:S7:S7-SP1",
"V2-SOL-04:S7:S7-SP2",
"V2-TERRA-01:s1_deepresearchbench:s1_dates",
"V2-TERRA-01:s1_deepresearchbench:s1_method",
"V2-TERRA-01:s1_deepresearchbench:s1_table",
"V2-TERRA-01:s2_reportbench:s2_collection",
"V2-TERRA-01:s2_reportbench:s2_method",
"V2-TERRA-01:s2_reportbench:s2_metric_table",
"V2-TERRA-01:s3_researcherbench:s3_definition",
"V2-TERRA-01:s3_researcherbench:s3_models_time",
"V2-TERRA-01:s3_researcherbench:s3_table",
"V2-TERRA-01:s4_cited_not_verified:s4_definition_table",
"V2-TERRA-01:s4_cited_not_verified:s4_depth",
"V2-TERRA-01:s5_urlhealth:s5_comparison",
"V2-TERRA-01:s5_urlhealth:s5_scope",
"V2-TERRA-01:s5_urlhealth:s5_table",
"V2-TERRA-02:s1_deepresearchbench:s1_fact_formula",
"V2-TERRA-02:s1_deepresearchbench:s1_fact_judge_validation",
"V2-TERRA-02:s1_deepresearchbench:s1_fact_method",
"V2-TERRA-02:s1_deepresearchbench:s1_fact_results",
"V2-TERRA-02:s2_liveresearchbench:s2_rubric_tree",
"V2-TERRA-02:s2_liveresearchbench:s2_table7",
"V2-TERRA-02:s2_liveresearchbench:s2_validation",
"V2-TERRA-02:s3_deeptrace:s3_corpus_and_retrieval",
"V2-TERRA-02:s3_deeptrace:s3_definition",
"V2-TERRA-02:s3_deeptrace:s3_judge_validation",
"V2-TERRA-02:s3_deeptrace:s3_table1",
"V2-TERRA-02:s4_researcherbench:s4_method",
"V2-TERRA-02:s4_researcherbench:s4_results",
"V2-TERRA-03:S1:S1-A",
"V2-TERRA-03:S1:S1-B",
"V2-TERRA-03:S1:S1-C",
"V2-TERRA-03:S1:S1-D",
"V2-TERRA-03:S1:S1-E",
"V2-TERRA-03:S1:S1-F",
"V2-TERRA-03:S2:S2-A",
"V2-TERRA-03:S2:S2-B",
"V2-TERRA-03:S2:S2-C",
"V2-TERRA-03:S2:S2-D",
"V2-TERRA-03:S3:S3-A",
"V2-TERRA-03:S3:S3-B",
"V2-TERRA-03:S3:S3-C",
"V2-TERRA-03:S3:S3-D",
"V2-TERRA-03:S4:S4-A",
"V2-TERRA-03:S4:S4-B",
"V2-TERRA-03:S4:S4-C",
"V2-TERRA-03:S5:S5-A",
"V2-TERRA-03:S5:S5-B",
"V2-TERRA-03:S6:S6-A",
"V2-TERRA-04:S1_deepresearchbench:S1_method",
"V2-TERRA-04:S1_deepresearchbench:S1_metric",
"V2-TERRA-04:S1_deepresearchbench:S1_table1",
"V2-TERRA-04:S1_deepresearchbench:S1_time",
"V2-TERRA-04:S2_researcherbench:S2_method",
"V2-TERRA-04:S2_researcherbench:S2_scope",
"V2-TERRA-04:S2_researcherbench:S2_table2",
"V2-TERRA-04:S2_researcherbench:S2_time",
"V2-TERRA-04:S3_reportbench:S3_limitations",
"V2-TERRA-04:S3_reportbench:S3_method",
"V2-TERRA-04:S3_reportbench:S3_table1",
"V2-TERRA-04:S3_reportbench:S3_time",
"V2-TERRA-04:S4_urlhealth:S4_comparison",
"V2-TERRA-04:S4_urlhealth:S4_limitations",
"V2-TERRA-04:S4_urlhealth:S4_method",
"V2-TERRA-04:S4_urlhealth:S4_table2",
"V2-TERRA-04:S5_cited_not_verified:S5_depth",
"V2-TERRA-04:S5_cited_not_verified:S5_limitations",
"V2-TERRA-04:S5_cited_not_verified:S5_method",
"V2-TERRA-04:S5_cited_not_verified:S5_scope",
"V2-TERRA-04:S5_cited_not_verified:S5_table1"
],
"rejected_claim_occurrence_ids": [
"V2-SOL-02:R5_deeptrace_audit",
"V2-SOL-03:r3_deepresearch_bench_fact",
"V2-SOL-03:r8_search_depth_ablation",
"V2-SOL-03:r9_verifier_calibration",
"V2-SOL-04:R1",
"V2-SOL-04:R2",
"V2-SOL-04:R3",
"V2-SOL-04:R4",
"V2-SOL-04:R8"
],
"reports": [
{
"answer_bytes": 40476,
"answer_path": "research/how-we-know/agent-citation-lineage/answers-v2/V2-SOL-01.json",
"answer_sha256": "17a7cf97b717d7b520026b9eb52a7da59ddb9b5f698c2aeadfab3f38c91008a0",
"prompt_sha256": "d321a9cec7b5fe419157c0623e18ff0020cb0080fbe6dd9ed4fd20a0b896f670",
"reported_model_identity": "unknown",
"requested_model_profile": "gpt-5.6-sol",
"retrieval_infrastructure": "unknown",
"run_id": "V2-SOL-01",
"status": "completed",
"trace_path": "research/how-we-know/agent-citation-lineage/traces-v2/V2-SOL-01.json"
},
{
"answer_bytes": 32376,
"answer_path": "research/how-we-know/agent-citation-lineage/answers-v2/V2-SOL-02.json",
"answer_sha256": "f72975b2b98532116ca9d2c510e2343df505bfce359ebfef595a0a35b77cb810",
"prompt_sha256": "d321a9cec7b5fe419157c0623e18ff0020cb0080fbe6dd9ed4fd20a0b896f670",
"reported_model_identity": "unknown",
"requested_model_profile": "gpt-5.6-sol",
"retrieval_infrastructure": "unknown",
"run_id": "V2-SOL-02",
"status": "completed",
"trace_path": "research/how-we-know/agent-citation-lineage/traces-v2/V2-SOL-02.json"
},
{
"answer_bytes": 38504,
"answer_path": "research/how-we-know/agent-citation-lineage/answers-v2/V2-SOL-03.json",
"answer_sha256": "782a60cc19d88217694cd672bac6f659f44fa0bd0f006f64a65ea77386d82dc7",
"prompt_sha256": "d321a9cec7b5fe419157c0623e18ff0020cb0080fbe6dd9ed4fd20a0b896f670",
"reported_model_identity": "unknown",
"requested_model_profile": "gpt-5.6-sol",
"retrieval_infrastructure": "unknown",
"run_id": "V2-SOL-03",
"status": "completed",
"trace_path": "research/how-we-know/agent-citation-lineage/traces-v2/V2-SOL-03.json"
},
{
"answer_bytes": 36500,
"answer_path": "research/how-we-know/agent-citation-lineage/answers-v2/V2-SOL-04.json",
"answer_sha256": "216552854069e32a8dc4986c57f59dc3b9b56d44f01d527167b3f311384a9261",
"prompt_sha256": "d321a9cec7b5fe419157c0623e18ff0020cb0080fbe6dd9ed4fd20a0b896f670",
"reported_model_identity": "unknown",
"requested_model_profile": "gpt-5.6-sol",
"retrieval_infrastructure": "credential-free public web search, open, click, and find operations; no logged-in browser session or account connector",
"run_id": "V2-SOL-04",
"status": "completed",
"trace_path": "research/how-we-know/agent-citation-lineage/traces-v2/V2-SOL-04.json"
},
{
"answer_bytes": 22598,
"answer_path": "research/how-we-know/agent-citation-lineage/answers-v2/V2-TERRA-01.json",
"answer_sha256": "e90c36b6f172fbe291d75ce1ea9451b1160515a2b7b4e0ead66aa1a76cc8b735",
"prompt_sha256": "d321a9cec7b5fe419157c0623e18ff0020cb0080fbe6dd9ed4fd20a0b896f670",
"reported_model_identity": "unknown",
"requested_model_profile": "gpt-5.6-terra",
"retrieval_infrastructure": "unknown",
"run_id": "V2-TERRA-01",
"status": "completed",
"trace_path": "research/how-we-know/agent-citation-lineage/traces-v2/V2-TERRA-01.json"
},
{
"answer_bytes": 20907,
"answer_path": "research/how-we-know/agent-citation-lineage/answers-v2/V2-TERRA-02.json",
"answer_sha256": "819d320162a09909a73df00a4e1ebbdd42e67794cb02e3f7c379f601e10a0192",
"prompt_sha256": "d321a9cec7b5fe419157c0623e18ff0020cb0080fbe6dd9ed4fd20a0b896f670",
"reported_model_identity": "unknown",
"requested_model_profile": "gpt-5.6-terra",
"retrieval_infrastructure": "unknown",
"run_id": "V2-TERRA-02",
"status": "completed",
"trace_path": "research/how-we-know/agent-citation-lineage/traces-v2/V2-TERRA-02.json"
},
{
"answer_bytes": 23183,
"answer_path": "research/how-we-know/agent-citation-lineage/answers-v2/V2-TERRA-03.json",
"answer_sha256": "e211dfe49066bb9f4ff9e59b91173d3b0f45b7b1e8ca5135c7435f3882d7e264",
"prompt_sha256": "d321a9cec7b5fe419157c0623e18ff0020cb0080fbe6dd9ed4fd20a0b896f670",
"reported_model_identity": "unknown",
"requested_model_profile": "gpt-5.6-terra",
"retrieval_infrastructure": "unknown",
"run_id": "V2-TERRA-03",
"status": "completed",
"trace_path": "research/how-we-know/agent-citation-lineage/traces-v2/V2-TERRA-03.json"
},
{
"answer_bytes": 30189,
"answer_path": "research/how-we-know/agent-citation-lineage/answers-v2/V2-TERRA-04.json",
"answer_sha256": "8deffa91b6bb52d1de4ef4965d0087eca3d3a8a96d971a5e89753c1928cc4ef4",
"prompt_sha256": "d321a9cec7b5fe419157c0623e18ff0020cb0080fbe6dd9ed4fd20a0b896f670",
"reported_model_identity": "unknown",
"requested_model_profile": "gpt-5.6-terra",
"retrieval_infrastructure": "unknown",
"run_id": "V2-TERRA-04",
"status": "completed",
"trace_path": "research/how-we-know/agent-citation-lineage/traces-v2/V2-TERRA-04.json"
}
],
"resolving_url_roots": [
"https://arxiv.org/abs/2604.03173",
"https://arxiv.org/abs/2605.06635",
"https://arxiv.org/abs/2607.08700",
"https://arxiv.org/html/2506.11763",
"https://arxiv.org/html/2506.11763v1",
"https://arxiv.org/html/2507.16280",
"https://arxiv.org/html/2508.15804",
"https://arxiv.org/html/2508.15804v1",
"https://arxiv.org/html/2509.04499",
"https://arxiv.org/html/2509.04499v1",
"https://arxiv.org/html/2604.03173",
"https://arxiv.org/html/2604.03173v1",
"https://arxiv.org/html/2605.06635",
"https://arxiv.org/html/2605.06635v1",
"https://arxiv.org/pdf/2507.16280",
"https://arxiv.org/pdf/2508.15804",
"https://arxiv.org/pdf/2510.14240v5",
"https://arxiv.org/pdf/2604.03173",
"https://arxiv.org/pdf/2605.06635",
"https://data.mendeley.com/datasets/3s73z9zf3c/1",
"https://data.mendeley.com/datasets/3s73z9zf3c/2",
"https://deepresearch-bench.github.io/static/papers/deepresearch-bench.pdf",
"https://github.com/Ayanami0730/deep_research_bench",
"https://pmc.ncbi.nlm.nih.gov/articles/PMC13109748/",
"https://proceedings.iclr.cc/paper_files/paper/2026/file/465f22be10e07b301c6ed58f0472f704-Paper-Conference.pdf",
"https://proceedings.iclr.cc/paper_files/paper/2026/file/ad08767706825033b99122332293033d-Paper-Conference.pdf",
"https://proceedings.iclr.cc/paper_files/paper/2026/hash/ad08767706825033b99122332293033d-Abstract-Conference.html"
],
"review_supplement_path": "research/how-we-know/agent-citation-lineage/review-supplement-spans-v1.json",
"review_supplement_sha256": "7daa7d56cf42f764448235074cf4e87abb81d4f3a99d16dae063b844dc6aa6c0",
"shared_capture_lineage": {
"automatic_independence_credit": 0,
"prompt_sha256": "d321a9cec7b5fe419157c0623e18ff0020cb0080fbe6dd9ed4fd20a0b896f670",
"reported_model_identity": "unknown",
"requested_profiles": [
"gpt-5.6-sol",
"gpt-5.6-terra"
],
"retrieval_infrastructure": "unknown"
},
"source_dependence_boundary": {
"automatic_independence_credit": 0,
"dimensions": [
"data",
"method",
"retrieval",
"source_work",
"edition",
"exact_span",
"derivation",
"upstream_citation"
],
"note": "Candidate warrant roots remain connected by declared or unresolved task, retrieval, judge-method, source, edition, span, and derivation lineage."
},
"source_work_ids": [
"work:citation-verifier-benchmark",
"work:cited-not-verified",
"work:deepresearch-bench-paper",
"work:deepresearch-bench-repository",
"work:deeptrace",
"work:keplinger-dermatology-audit",
"work:keplinger-supplement",
"work:liveresearchbench",
"work:reportbench",
"work:researcherbench",
"work:url-health"
],
"unresolved_citations": [
{
"citation_occurrence_id": "V2-SOL-01:s1_keplinger",
"correction_ids": [],
"edition_id": "edition:keplinger-vor-2025",
"license": "CC BY-NC 4.0",
"license_treatment": "quote-minimal attributed spans; no full-text redistribution",
"raw_source_id": "s1_keplinger",
"raw_title": "Assessment of Deep Research for dermatology literature reviews: Deep concern over the hype",
"readback": {
"captured_bytes": 111993,
"captured_sha256": "abbaa9f22035230d13152a68aaeecd8e83e40867c9dae4acd1d547e36c866793",
"edition_id": "edition:keplinger-vor-2025",
"http_status": 200,
"media_type": "text/html; charset=utf-8",
"redirect_chain": [],
"requested_url": "https://pmc.ncbi.nlm.nih.gov/articles/PMC13109748/",
"resolved_url": "https://pmc.ncbi.nlm.nih.gov/articles/PMC13109748/",
"retrieval_status": "retrieved"
},
"requested_url": "https://pmc.ncbi.nlm.nih.gov/articles/PMC13109748/",
"resolution_status": "unresolved",
"run_id": "V2-SOL-01",
"source_work_id": "work:keplinger-dermatology-audit",
"span_occurrence_ids": [
"V2-SOL-01:s1_keplinger:sp1a",
"V2-SOL-01:s1_keplinger:sp1b",
"V2-SOL-01:s1_keplinger:sp1c"
]
},
{
"citation_occurrence_id": "V2-SOL-01:s1a_keplinger_data",
"correction_ids": [
"correction:mendeley-file-readback"
],
"edition_id": "edition:keplinger-supplement-v2",
"license": "CC BY 4.0",
"license_treatment": "metadata and quote-minimal landing-page spans; file API required authentication and was not used",
"raw_source_id": "s1a_keplinger_data",
"raw_title": "Supplementary materials of the article: Assessment of Deep Research for Dermatology Literature Reviews: Deep Concern Over the Hype",
"readback": {
"captured_bytes": 116951,
"captured_sha256": "0a2c4c6ed54dd7e2e0ca6cc07aa19b18fa1439e506c2310db778094282d2cebc",
"edition_id": "edition:keplinger-supplement-v2",
"http_status": 200,
"media_type": "text/html; charset=utf-8",
"redirect_chain": [],
"requested_url": "https://data.mendeley.com/datasets/3s73z9zf3c/2",
"resolved_url": "https://data.mendeley.com/datasets/3s73z9zf3c/2",
"retrieval_status": "retrieved"
},
"requested_url": "https://data.mendeley.com/datasets/3s73z9zf3c/2",
"resolution_status": "unresolved",
"run_id": "V2-SOL-01",
"source_work_id": "work:keplinger-supplement",
"span_occurrence_ids": [
"V2-SOL-01:s1a_keplinger_data:sp1d"
]
},
{
"citation_occurrence_id": "V2-SOL-01:s2_drbench",
"correction_ids": [],
"edition_id": "edition:drbench-iclr-2026",
"license": "unknown",
"license_treatment": "metadata, digest, and quote-minimal attributed spans only",
"raw_source_id": "s2_drbench",
"raw_title": "DeepResearch Bench: A Comprehensive Benchmark for Deep Research Agents",
"readback": {
"captured_bytes": 2665857,
"captured_sha256": "8f80ce247f7cc355bb6773f36037e01cc3ba2e2082c77381eaadf0d2b92b021c",
"edition_id": "edition:drbench-iclr-2026",
"http_status": 200,
"media_type": "application/pdf",
"redirect_chain": [],
"requested_url": "https://proceedings.iclr.cc/paper_files/paper/2026/file/465f22be10e07b301c6ed58f0472f704-Paper-Conference.pdf",
"resolved_url": "https://proceedings.iclr.cc/paper_files/paper/2026/file/465f22be10e07b301c6ed58f0472f704-Paper-Conference.pdf",
"retrieval_status": "retrieved"
},
"requested_url": "https://proceedings.iclr.cc/paper_files/paper/2026/file/465f22be10e07b301c6ed58f0472f704-Paper-Conference.pdf",
"resolution_status": "unresolved",
"run_id": "V2-SOL-01",
"source_work_id": "work:deepresearch-bench-paper",
"span_occurrence_ids": [
"V2-SOL-01:s2_drbench:sp2a",
"V2-SOL-01:s2_drbench:sp2b",
"V2-SOL-01:s2_drbench:sp2c",
"V2-SOL-01:s2_drbench:sp2d"
]
},
{
"citation_occurrence_id": "V2-SOL-01:s3_deeptrace",
"correction_ids": [],
"edition_id": "edition:deeptrace-iclr-2026",
"license": "unknown",
"license_treatment": "metadata, digest, and quote-minimal attributed spans only",
"raw_source_id": "s3_deeptrace",
"raw_title": "DeepTRACE: Auditing Deep Research AI Systems for Tracking Reliability Across Citations and Evidence",
"readback": {
"captured_bytes": 2982839,
"captured_sha256": "dea4981c1066d0240a005f603b3b14419c7e32beb5191b3df847ceb26af3d6b6",
"edition_id": "edition:deeptrace-iclr-2026",
"http_status": 200,
"media_type": "application/pdf",
"redirect_chain": [],
"requested_url": "https://proceedings.iclr.cc/paper_files/paper/2026/file/ad08767706825033b99122332293033d-Paper-Conference.pdf",
"resolved_url": "https://proceedings.iclr.cc/paper_files/paper/2026/file/ad08767706825033b99122332293033d-Paper-Conference.pdf",
"retrieval_status": "retrieved"
},
"requested_url": "https://proceedings.iclr.cc/paper_files/paper/2026/file/ad08767706825033b99122332293033d-Paper-Conference.pdf",
"resolution_status": "unresolved",
"run_id": "V2-SOL-01",
"source_work_id": "work:deeptrace",
"span_occurrence_ids": [
"V2-SOL-01:s3_deeptrace:sp3a",
"V2-SOL-01:s3_deeptrace:sp3b",
"V2-SOL-01:s3_deeptrace:sp3c",
"V2-SOL-01:s3_deeptrace:sp3d"
]
},
{
"citation_occurrence_id": "V2-SOL-01:s4_cited_not_verified",
"correction_ids": [],
"edition_id": "edition:cnv-arxiv-v1",
"license": "CC BY 4.0",
"license_treatment": "quote-minimal attributed spans",
"raw_source_id": "s4_cited_not_verified",
"raw_title": "Cited but Not Verified: Parsing and Evaluating Source Attribution in LLM Deep Research Agents",
"readback": {
"captured_bytes": 43848,
"captured_sha256": "d1b6d476b3e81460dbc8821757775a8fa589e0579b77167ef0f33d35cb819cb5",
"edition_id": "edition:cnv-arxiv-v1",
"http_status": 200,
"media_type": "text/html; charset=utf-8",
"redirect_chain": [],
"requested_url": "https://arxiv.org/abs/2605.06635",
"resolved_url": "https://arxiv.org/abs/2605.06635",
"retrieval_status": "retrieved"
},
"requested_url": "https://arxiv.org/abs/2605.06635",
"resolution_status": "unresolved",
"run_id": "V2-SOL-01",
"source_work_id": "work:cited-not-verified",
"span_occurrence_ids": [
"V2-SOL-01:s4_cited_not_verified:sp4a",
"V2-SOL-01:s4_cited_not_verified:sp4b",
"V2-SOL-01:s4_cited_not_verified:sp4c"
]
},
{
"citation_occurrence_id": "V2-SOL-01:s5_url_health",
"correction_ids": [],
"edition_id": "edition:url-health-arxiv-v1",
"license": "CC0 1.0",
"license_treatment": "quote-minimal attributed spans",
"raw_source_id": "s5_url_health",
"raw_title": "Detecting and Correcting Reference Hallucinations in Commercial LLMs and Deep Research Agents",
"readback": {
"captured_bytes": 42408,
"captured_sha256": "3976893e82f91cc9e7826c0d2d08c6caab40778134dbab09ebbfe1f6f63f5397",
"edition_id": "edition:url-health-arxiv-v1",
"http_status": 200,
"media_type": "text/html; charset=utf-8",
"redirect_chain": [],
"requested_url": "https://arxiv.org/abs/2604.03173",
"resolved_url": "https://arxiv.org/abs/2604.03173",
"retrieval_status": "retrieved"
},
"requested_url": "https://arxiv.org/abs/2604.03173",
"resolution_status": "unresolved",
"run_id": "V2-SOL-01",
"source_work_id": "work:url-health",
"span_occurrence_ids": [
"V2-SOL-01:s5_url_health:sp5a",
"V2-SOL-01:s5_url_health:sp5b",
"V2-SOL-01:s5_url_health:sp5c"
]
},
{
"citation_occurrence_id": "V2-SOL-02:S1",
"correction_ids": [],
"edition_id": "edition:url-health-arxiv-v1",
"license": "CC0 1.0",
"license_treatment": "quote-minimal attributed spans",
"raw_source_id": "S1",
"raw_title": "Detecting and Correcting Reference Hallucinations in Commercial LLMs and Deep Research Agents",
"readback": {
"captured_bytes": 302498,
"captured_sha256": "332e5b5cb4b0ee7065b1bbc30436dfdfeafca8130fb87e32b0020026d8bbafe1",
"edition_id": "edition:url-health-arxiv-v1",
"http_status": 200,
"media_type": "text/html; charset=utf-8",
"redirect_chain": [],
"requested_url": "https://arxiv.org/html/2604.03173v1",
"resolved_url": "https://arxiv.org/html/2604.03173v1",
"retrieval_status": "retrieved"
},
"requested_url": "https://arxiv.org/html/2604.03173v1",
"resolution_status": "unresolved",
"run_id": "V2-SOL-02",
"source_work_id": "work:url-health",
"span_occurrence_ids": [
"V2-SOL-02:S1:S1_SPAN_1",
"V2-SOL-02:S1:S1_SPAN_2"
]
},
{
"citation_occurrence_id": "V2-SOL-02:S3",
"correction_ids": [
"correction:mendeley-file-readback"
],
"edition_id": "edition:keplinger-supplement-v2",
"license": "CC BY 4.0",
"license_treatment": "metadata and quote-minimal landing-page spans; file API required authentication and was not used",
"raw_source_id": "S3",
"raw_title": "Supplementary materials of the article: Assessment of Deep Research for Dermatology Literature Reviews: Deep Concern Over the Hype",
"readback": {
"captured_bytes": 116951,
"captured_sha256": "0a2c4c6ed54dd7e2e0ca6cc07aa19b18fa1439e506c2310db778094282d2cebc",
"edition_id": "edition:keplinger-supplement-v2",
"http_status": 200,
"media_type": "text/html; charset=utf-8",
"redirect_chain": [],
"requested_url": "https://data.mendeley.com/datasets/3s73z9zf3c/2",
"resolved_url": "https://data.mendeley.com/datasets/3s73z9zf3c/2",
"retrieval_status": "retrieved"
},
"requested_url": "https://data.mendeley.com/datasets/3s73z9zf3c/2",
"resolution_status": "unresolved",
"run_id": "V2-SOL-02",
"source_work_id": "work:keplinger-supplement",
"span_occurrence_ids": [
"V2-SOL-02:S3:S3_SPAN_1",
"V2-SOL-02:S3:S3_SPAN_2"
]
},
{
"citation_occurrence_id": "V2-SOL-02:S4",
"correction_ids": [],
"edition_id": "edition:drbench-iclr-2026",
"license": "unknown",
"license_treatment": "metadata, digest, and quote-minimal attributed spans only",
"raw_source_id": "S4",
"raw_title": "DeepResearch Bench: A Comprehensive Benchmark for Deep Research Agents",
"readback": {
"captured_bytes": 12692,
"captured_sha256": "a2d811e0d24af89d9bd4b19cc9e8fbc474c67ef3827648550a6560eca2a5e332",
"edition_id": "edition:drbench-iclr-2026",
"http_status": 403,
"media_type": "text/html",
"redirect_chain": [],
"requested_url": "https://openreview.net/pdf?id=hQ0K2Hhq7H",
"resolved_url": "https://openreview.net/pdf?id=hQ0K2Hhq7H",
"retrieval_status": "inaccessible"
},
"requested_url": "https://openreview.net/pdf?id=hQ0K2Hhq7H",
"resolution_status": "unresolved",
"run_id": "V2-SOL-02",
"source_work_id": "work:deepresearch-bench-paper",
"span_occurrence_ids": [
"V2-SOL-02:S4:S4_SPAN_1",
"V2-SOL-02:S4:S4_SPAN_2"
]
},
{
"citation_occurrence_id": "V2-SOL-02:S5",
"correction_ids": [],
"edition_id": "edition:reportbench-arxiv-v1",
"license": "CC BY 4.0",
"license_treatment": "quote-minimal attributed spans",
"raw_source_id": "S5",
"raw_title": "ReportBench: Evaluating Deep Research Agents via Academic Survey Tasks",
"readback": {
"captured_bytes": 157386,
"captured_sha256": "055e568189a402dd510c7c84be60f59465d9a001808936bc25cb0026ac25d267",
"edition_id": "edition:reportbench-arxiv-v1",
"http_status": 200,
"media_type": "text/html; charset=utf-8",
"redirect_chain": [],
"requested_url": "https://arxiv.org/html/2508.15804v1",
"resolved_url": "https://arxiv.org/html/2508.15804v1",
"retrieval_status": "retrieved"
},
"requested_url": "https://arxiv.org/html/2508.15804v1",
"resolution_status": "unresolved",
"run_id": "V2-SOL-02",
"source_work_id": "work:reportbench",
"span_occurrence_ids": [
"V2-SOL-02:S5:S5_SPAN_1",
"V2-SOL-02:S5:S5_SPAN_2"
]
},
{
"citation_occurrence_id": "V2-SOL-02:S7",
"correction_ids": [],
"edition_id": "edition:cnv-arxiv-v1",
"license": "CC BY 4.0",
"license_treatment": "quote-minimal attributed spans",
"raw_source_id": "S7",
"raw_title": "Cited but Not Verified: Parsing and Evaluating Source Attribution in LLM Deep Research Agents",
"readback": {
"captured_bytes": 138385,
"captured_sha256": "7c5e3c33f3122b07d176e6679cd6762a95a6babaf4a353686f019f905b5526ca",
"edition_id": "edition:cnv-arxiv-v1",
"http_status": 200,
"media_type": "text/html; charset=utf-8",
"redirect_chain": [],
"requested_url": "https://arxiv.org/html/2605.06635v1",
"resolved_url": "https://arxiv.org/html/2605.06635v1",
"retrieval_status": "retrieved"
},
"requested_url": "https://arxiv.org/html/2605.06635v1",
"resolution_status": "unresolved",
"run_id": "V2-SOL-02",
"source_work_id": "work:cited-not-verified",
"span_occurrence_ids": [
"V2-SOL-02:S7:S7_SPAN_1",
"V2-SOL-02:S7:S7_SPAN_2"
]
},
{
"citation_occurrence_id": "V2-SOL-03:s1",
"correction_ids": [],
"edition_id": "edition:keplinger-vor-2025",
"license": "CC BY-NC 4.0",
"license_treatment": "quote-minimal attributed spans; no full-text redistribution",
"raw_source_id": "s1",
"raw_title": "Assessment of Deep Research for dermatology literature reviews: Deep concern over the hype",
"readback": {
"captured_bytes": 5498,
"captured_sha256": "d74c1891ae09a2b89121f2a60ac837c89761a8086759e45db214ca68c1905cea",
"edition_id": "edition:keplinger-vor-2025",
"http_status": 403,
"media_type": "text/html; charset=UTF-8",
"redirect_chain": [],
"requested_url": "https://onlinelibrary.wiley.com/doi/10.1111/jdv.70035",
"resolved_url": "https://onlinelibrary.wiley.com/doi/10.1111/jdv.70035",
"retrieval_status": "inaccessible"
},
"requested_url": "https://onlinelibrary.wiley.com/doi/10.1111/jdv.70035",
"resolution_status": "unresolved",
"run_id": "V2-SOL-03",
"source_work_id": "work:keplinger-dermatology-audit",
"span_occurrence_ids": [
"V2-SOL-03:s1:s1_span1"
]
},
{
"citation_occurrence_id": "V2-SOL-03:s3",
"correction_ids": [],
"edition_id": "edition:deeptrace-arxiv-v1",
"license": "arXiv non-exclusive distribution license",
"license_treatment": "metadata, digest, and quote-minimal attributed spans only",
"raw_source_id": "s3",
"raw_title": "DeepTRACE: Auditing Deep Research AI Systems for Tracking Reliability Across Citations and Evidence",
"readback": {
"captured_bytes": 364455,
"captured_sha256": "0c28edecaedd882584e985caac1e14f4a03dd59e939e17ae755a3c5e07ae42b0",
"edition_id": "edition:deeptrace-arxiv-v1",
"http_status": 200,
"media_type": "text/html; charset=utf-8",
"redirect_chain": [],
"requested_url": "https://arxiv.org/html/2509.04499",
"resolved_url": "https://arxiv.org/html/2509.04499",
"retrieval_status": "retrieved"
},
"requested_url": "https://arxiv.org/html/2509.04499",
"resolution_status": "unresolved",
"run_id": "V2-SOL-03",
"source_work_id": "work:deeptrace",
"span_occurrence_ids": [
"V2-SOL-03:s3:s3_span1"
]
},
{
"citation_occurrence_id": "V2-SOL-03:s4",
"correction_ids": [],
"edition_id": "edition:url-health-arxiv-v1",
"license": "CC0 1.0",
"license_treatment": "quote-minimal attributed spans",
"raw_source_id": "s4",
"raw_title": "Detecting and Correcting Reference Hallucinations in Commercial LLMs and Deep Research Agents",
"readback": {
"captured_bytes": 302498,
"captured_sha256": "332e5b5cb4b0ee7065b1bbc30436dfdfeafca8130fb87e32b0020026d8bbafe1",
"edition_id": "edition:url-health-arxiv-v1",
"http_status": 200,
"media_type": "text/html; charset=utf-8",
"redirect_chain": [],
"requested_url": "https://arxiv.org/html/2604.03173",
"resolved_url": "https://arxiv.org/html/2604.03173",
"retrieval_status": "retrieved"
},
"requested_url": "https://arxiv.org/html/2604.03173",
"resolution_status": "unresolved",
"run_id": "V2-SOL-03",
"source_work_id": "work:url-health",
"span_occurrence_ids": [
"V2-SOL-03:s4:s4_span1",
"V2-SOL-03:s4:s4_span2",
"V2-SOL-03:s4:s4_span3"
]
},
{
"citation_occurrence_id": "V2-SOL-04:S6",
"correction_ids": [
"correction:mendeley-file-readback"
],
"edition_id": "edition:keplinger-supplement-v1",
"license": "CC BY 4.0",
"license_treatment": "metadata and quote-minimal landing-page spans",
"raw_source_id": "S6",
"raw_title": "Supplementary materials of the article: Assessment of Deep Research for Dermatology Literature Reviews: Deep Concern Over the Hype",
"readback": {
"captured_bytes": 115612,
"captured_sha256": "da6308937337b2dfd3f97dc62f817e4a7ec362329ede0fce4c9bcdfd99ad93b1",
"edition_id": "edition:keplinger-supplement-v1",
"http_status": 200,
"media_type": "text/html; charset=utf-8",
"redirect_chain": [],
"requested_url": "https://data.mendeley.com/datasets/3s73z9zf3c/1",
"resolved_url": "https://data.mendeley.com/datasets/3s73z9zf3c/1",
"retrieval_status": "retrieved"
},
"requested_url": "https://data.mendeley.com/datasets/3s73z9zf3c/1",
"resolution_status": "unresolved",
"run_id": "V2-SOL-04",
"source_work_id": "work:keplinger-supplement",
"span_occurrence_ids": [
"V2-SOL-04:S6:S6-SP1"
]
},
{
"citation_occurrence_id": "V2-TERRA-01:s1_deepresearchbench",
"correction_ids": [],
"edition_id": "edition:drbench-arxiv-v1",
"license": "CC BY 4.0",
"license_treatment": "quote-minimal attributed spans",
"raw_source_id": "s1_deepresearchbench",
"raw_title": "DeepResearch Bench: A Comprehensive Benchmark for Deep Research Agents",
"readback": {
"captured_bytes": 3678399,
"captured_sha256": "8fbf30398f5e62f8839f0c9c8609bbb9e3cd0b57ae27d4bf33cb5db2007d1118",
"edition_id": "edition:drbench-arxiv-v1",
"http_status": 200,
"media_type": "application/pdf",
"redirect_chain": [],
"requested_url": "https://deepresearch-bench.github.io/static/papers/deepresearch-bench.pdf",
"resolved_url": "https://deepresearch-bench.github.io/static/papers/deepresearch-bench.pdf",
"retrieval_status": "retrieved"
},
"requested_url": "https://deepresearch-bench.github.io/static/papers/deepresearch-bench.pdf",
"resolution_status": "unresolved",
"run_id": "V2-TERRA-01",
"source_work_id": "work:deepresearch-bench-paper",
"span_occurrence_ids": [
"V2-TERRA-01:s1_deepresearchbench:s1_method",
"V2-TERRA-01:s1_deepresearchbench:s1_table",
"V2-TERRA-01:s1_deepresearchbench:s1_dates"
]
},
{
"citation_occurrence_id": "V2-TERRA-01:s2_reportbench",
"correction_ids": [],
"edition_id": "edition:reportbench-arxiv-v1",
"license": "CC BY 4.0",
"license_treatment": "quote-minimal attributed spans",
"raw_source_id": "s2_reportbench",
"raw_title": "ReportBench: Evaluating Deep Research Agents via Academic Survey Tasks",
"readback": {
"captured_bytes": 729607,
"captured_sha256": "90730ad75011d460305caf45a5be33b1b5eb4126f7e7efc1506f63beda1f91d3",
"edition_id": "edition:reportbench-arxiv-v1",
"http_status": 200,
"media_type": "application/pdf",
"redirect_chain": [],
"requested_url": "https://arxiv.org/pdf/2508.15804",
"resolved_url": "https://arxiv.org/pdf/2508.15804",
"retrieval_status": "retrieved"
},
"requested_url": "https://arxiv.org/pdf/2508.15804",
"resolution_status": "unresolved",
"run_id": "V2-TERRA-01",
"source_work_id": "work:reportbench",
"span_occurrence_ids": [
"V2-TERRA-01:s2_reportbench:s2_method",
"V2-TERRA-01:s2_reportbench:s2_metric_table",
"V2-TERRA-01:s2_reportbench:s2_collection"
]
},
{
"citation_occurrence_id": "V2-TERRA-01:s3_researcherbench",
"correction_ids": [],
"edition_id": "edition:researcherbench-arxiv-v1",
"license": "arXiv non-exclusive distribution license",
"license_treatment": "metadata, digest, and quote-minimal attributed spans only",
"raw_source_id": "s3_researcherbench",
"raw_title": "ResearcherBench: Evaluating Deep AI Research Systems on the Frontiers of Scientific Inquiry",
"readback": {
"captured_bytes": 1566914,
"captured_sha256": "1571435270d3e781ab8d5254913af00a7dcf0efcb99ec8e6d3db87a7b5c92f02",
"edition_id": "edition:researcherbench-arxiv-v1",
"http_status": 200,
"media_type": "application/pdf",
"redirect_chain": [],
"requested_url": "https://arxiv.org/pdf/2507.16280",
"resolved_url": "https://arxiv.org/pdf/2507.16280",
"retrieval_status": "retrieved"
},
"requested_url": "https://arxiv.org/pdf/2507.16280",
"resolution_status": "unresolved",
"run_id": "V2-TERRA-01",
"source_work_id": "work:researcherbench",
"span_occurrence_ids": [
"V2-TERRA-01:s3_researcherbench:s3_definition",
"V2-TERRA-01:s3_researcherbench:s3_models_time",
"V2-TERRA-01:s3_researcherbench:s3_table"
]
},
{
"citation_occurrence_id": "V2-TERRA-01:s4_cited_not_verified",
"correction_ids": [],
"edition_id": "edition:cnv-arxiv-v1",
"license": "CC BY 4.0",
"license_treatment": "quote-minimal attributed spans",
"raw_source_id": "s4_cited_not_verified",
"raw_title": "Cited but Not Verified: Parsing and Evaluating Source Attribution in LLM Deep Research Agents",
"readback": {
"captured_bytes": 1424589,
"captured_sha256": "db5b7b7e3d9ce9fc6d6713b60f65f0d81b07090ded2a28be7f541d1c290230b5",
"edition_id": "edition:cnv-arxiv-v1",
"http_status": 200,
"media_type": "application/pdf",
"redirect_chain": [],
"requested_url": "https://arxiv.org/pdf/2605.06635",
"resolved_url": "https://arxiv.org/pdf/2605.06635",
"retrieval_status": "retrieved"
},
"requested_url": "https://arxiv.org/pdf/2605.06635",
"resolution_status": "unresolved",
"run_id": "V2-TERRA-01",
"source_work_id": "work:cited-not-verified",
"span_occurrence_ids": [
"V2-TERRA-01:s4_cited_not_verified:s4_definition_table",
"V2-TERRA-01:s4_cited_not_verified:s4_depth"
]
},
{
"citation_occurrence_id": "V2-TERRA-01:s5_urlhealth",
"correction_ids": [],
"edition_id": "edition:url-health-arxiv-v1",
"license": "CC0 1.0",
"license_treatment": "quote-minimal attributed spans",
"raw_source_id": "s5_urlhealth",
"raw_title": "Detecting and Correcting Reference Hallucinations in Commercial LLMs and Deep Research Agents",
"readback": {
"captured_bytes": 416944,
"captured_sha256": "e8145a9f62f2a2bf0c5f3e2607160c56e010d02da965b5f65dbfa8f07ddde06c",
"edition_id": "edition:url-health-arxiv-v1",
"http_status": 200,
"media_type": "application/pdf",
"redirect_chain": [],
"requested_url": "https://arxiv.org/pdf/2604.03173",
"resolved_url": "https://arxiv.org/pdf/2604.03173",
"retrieval_status": "retrieved"
},
"requested_url": "https://arxiv.org/pdf/2604.03173",
"resolution_status": "unresolved",
"run_id": "V2-TERRA-01",
"source_work_id": "work:url-health",
"span_occurrence_ids": [
"V2-TERRA-01:s5_urlhealth:s5_scope",
"V2-TERRA-01:s5_urlhealth:s5_table",
"V2-TERRA-01:s5_urlhealth:s5_comparison"
]
},
{
"citation_occurrence_id": "V2-TERRA-02:s1_deepresearchbench",
"correction_ids": [],
"edition_id": "edition:drbench-arxiv-v1",
"license": "CC BY 4.0",
"license_treatment": "quote-minimal attributed spans",
"raw_source_id": "s1_deepresearchbench",
"raw_title": "DeepResearch Bench: A Comprehensive Benchmark for Deep Research Agents",
"readback": {
"captured_bytes": 3678399,
"captured_sha256": "8fbf30398f5e62f8839f0c9c8609bbb9e3cd0b57ae27d4bf33cb5db2007d1118",
"edition_id": "edition:drbench-arxiv-v1",
"http_status": 200,
"media_type": "application/pdf",
"redirect_chain": [],
"requested_url": "https://deepresearch-bench.github.io/static/papers/deepresearch-bench.pdf",
"resolved_url": "https://deepresearch-bench.github.io/static/papers/deepresearch-bench.pdf",
"retrieval_status": "retrieved"
},
"requested_url": "https://deepresearch-bench.github.io/static/papers/deepresearch-bench.pdf",
"resolution_status": "unresolved",
"run_id": "V2-TERRA-02",
"source_work_id": "work:deepresearch-bench-paper",
"span_occurrence_ids": [
"V2-TERRA-02:s1_deepresearchbench:s1_fact_method",
"V2-TERRA-02:s1_deepresearchbench:s1_fact_results",
"V2-TERRA-02:s1_deepresearchbench:s1_fact_formula",
"V2-TERRA-02:s1_deepresearchbench:s1_fact_judge_validation"
]
},
{
"citation_occurrence_id": "V2-TERRA-02:s2_liveresearchbench",
"correction_ids": [
"correction:liveresearchbench-license"
],
"edition_id": "edition:liveresearchbench-arxiv-v5",
"license": "CC BY-NC-SA 4.0",
"license_treatment": "quote-minimal attributed spans; raw CC BY claim is corrected in review records",
"raw_source_id": "s2_liveresearchbench",
"raw_title": "LiveResearchBench: A Live Benchmark for User-Centric Deep Research in the Wild",
"readback": {
"captured_bytes": 13057891,
"captured_sha256": "579b9728b76cfd242e9c94d9ff2985e196bbc72b5a741030e4f308ede04a4f69",
"edition_id": "edition:liveresearchbench-arxiv-v5",
"http_status": 200,
"media_type": "application/pdf",
"redirect_chain": [],
"requested_url": "https://arxiv.org/pdf/2510.14240v5",
"resolved_url": "https://arxiv.org/pdf/2510.14240v5",
"retrieval_status": "retrieved"
},
"requested_url": "https://arxiv.org/pdf/2510.14240v5",
"resolution_status": "unresolved",
"run_id": "V2-TERRA-02",
"source_work_id": "work:liveresearchbench",
"span_occurrence_ids": [
"V2-TERRA-02:s2_liveresearchbench:s2_rubric_tree",
"V2-TERRA-02:s2_liveresearchbench:s2_table7",
"V2-TERRA-02:s2_liveresearchbench:s2_validation"
]
},
{
"citation_occurrence_id": "V2-TERRA-02:s3_deeptrace",
"correction_ids": [],
"edition_id": "edition:deeptrace-iclr-2026",
"license": "unknown",
"license_treatment": "metadata, digest, and quote-minimal attributed spans only",
"raw_source_id": "s3_deeptrace",
"raw_title": "DeepTRACE: Auditing Deep Research AI Systems for Tracking Reliability Across Citations and Evidence",
"readback": {
"captured_bytes": 2982839,
"captured_sha256": "dea4981c1066d0240a005f603b3b14419c7e32beb5191b3df847ceb26af3d6b6",
"edition_id": "edition:deeptrace-iclr-2026",
"http_status": 200,
"media_type": "application/pdf",
"redirect_chain": [],
"requested_url": "https://proceedings.iclr.cc/paper_files/paper/2026/file/ad08767706825033b99122332293033d-Paper-Conference.pdf",
"resolved_url": "https://proceedings.iclr.cc/paper_files/paper/2026/file/ad08767706825033b99122332293033d-Paper-Conference.pdf",
"retrieval_status": "retrieved"
},
"requested_url": "https://proceedings.iclr.cc/paper_files/paper/2026/file/ad08767706825033b99122332293033d-Paper-Conference.pdf",
"resolution_status": "unresolved",
"run_id": "V2-TERRA-02",
"source_work_id": "work:deeptrace",
"span_occurrence_ids": [
"V2-TERRA-02:s3_deeptrace:s3_definition",
"V2-TERRA-02:s3_deeptrace:s3_table1",
"V2-TERRA-02:s3_deeptrace:s3_corpus_and_retrieval",
"V2-TERRA-02:s3_deeptrace:s3_judge_validation"
]
},
{
"citation_occurrence_id": "V2-TERRA-02:s4_researcherbench",
"correction_ids": [],
"edition_id": "edition:researcherbench-arxiv-v1",
"license": "arXiv non-exclusive distribution license",
"license_treatment": "metadata, digest, and quote-minimal attributed spans only",
"raw_source_id": "s4_researcherbench",
"raw_title": "ResearcherBench: Evaluating Deep AI Research Systems on the Frontiers of Scientific Inquiry",
"readback": {
"captured_bytes": 280402,
"captured_sha256": "c080d7304274d70a39651f49297cc65910ca833d7ee6a3de3052370733e44ee3",
"edition_id": "edition:researcherbench-arxiv-v1",
"http_status": 200,
"media_type": "text/html; charset=utf-8",
"redirect_chain": [],
"requested_url": "https://arxiv.org/html/2507.16280",
"resolved_url": "https://arxiv.org/html/2507.16280",
"retrieval_status": "retrieved"
},
"requested_url": "https://arxiv.org/html/2507.16280",
"resolution_status": "unresolved",
"run_id": "V2-TERRA-02",
"source_work_id": "work:researcherbench",
"span_occurrence_ids": [
"V2-TERRA-02:s4_researcherbench:s4_method",
"V2-TERRA-02:s4_researcherbench:s4_results"
]
},
{
"citation_occurrence_id": "V2-TERRA-03:S1",
"correction_ids": [],
"edition_id": "edition:cnv-arxiv-v1",
"license": "CC BY 4.0",
"license_treatment": "quote-minimal attributed spans",
"raw_source_id": "S1",
"raw_title": "Cited but Not Verified: Parsing and Evaluating Source Attribution in LLM Deep Research Agents",
"readback": {
"captured_bytes": 138385,
"captured_sha256": "7c5e3c33f3122b07d176e6679cd6762a95a6babaf4a353686f019f905b5526ca",
"edition_id": "edition:cnv-arxiv-v1",
"http_status": 200,
"media_type": "text/html; charset=utf-8",
"redirect_chain": [],
"requested_url": "https://arxiv.org/html/2605.06635",
"resolved_url": "https://arxiv.org/html/2605.06635",
"retrieval_status": "retrieved"
},
"requested_url": "https://arxiv.org/html/2605.06635",
"resolution_status": "unresolved",
"run_id": "V2-TERRA-03",
"source_work_id": "work:cited-not-verified",
"span_occurrence_ids": [
"V2-TERRA-03:S1:S1-A",
"V2-TERRA-03:S1:S1-B",
"V2-TERRA-03:S1:S1-C",
"V2-TERRA-03:S1:S1-D",
"V2-TERRA-03:S1:S1-E",
"V2-TERRA-03:S1:S1-F"
]
},
{
"citation_occurrence_id": "V2-TERRA-03:S2",
"correction_ids": [],
"edition_id": "edition:drbench-arxiv-v1",
"license": "CC BY 4.0",
"license_treatment": "quote-minimal attributed spans",
"raw_source_id": "S2",
"raw_title": "DeepResearch Bench: A Comprehensive Benchmark for Deep Research Agents",
"readback": {
"captured_bytes": 352002,
"captured_sha256": "9aa2894dbeaac30b23e7ffc8107a7f53b6e3855c8511838551de5bd7a422cc42",
"edition_id": "edition:drbench-arxiv-v1",
"http_status": 200,
"media_type": "text/html; charset=utf-8",
"redirect_chain": [],
"requested_url": "https://arxiv.org/html/2506.11763",
"resolved_url": "https://arxiv.org/html/2506.11763",
"retrieval_status": "retrieved"
},
"requested_url": "https://arxiv.org/html/2506.11763",
"resolution_status": "unresolved",
"run_id": "V2-TERRA-03",
"source_work_id": "work:deepresearch-bench-paper",
"span_occurrence_ids": [
"V2-TERRA-03:S2:S2-A",
"V2-TERRA-03:S2:S2-B",
"V2-TERRA-03:S2:S2-C",
"V2-TERRA-03:S2:S2-D"
]
},
{
"citation_occurrence_id": "V2-TERRA-03:S3",
"correction_ids": [],
"edition_id": "edition:reportbench-arxiv-v1",
"license": "CC BY 4.0",
"license_treatment": "quote-minimal attributed spans",
"raw_source_id": "S3",
"raw_title": "ReportBench: Evaluating Deep Research Agents via Academic Survey Tasks",
"readback": {
"captured_bytes": 157386,
"captured_sha256": "055e568189a402dd510c7c84be60f59465d9a001808936bc25cb0026ac25d267",
"edition_id": "edition:reportbench-arxiv-v1",
"http_status": 200,
"media_type": "text/html; charset=utf-8",
"redirect_chain": [],
"requested_url": "https://arxiv.org/html/2508.15804",
"resolved_url": "https://arxiv.org/html/2508.15804",
"retrieval_status": "retrieved"
},
"requested_url": "https://arxiv.org/html/2508.15804",
"resolution_status": "unresolved",
"run_id": "V2-TERRA-03",
"source_work_id": "work:reportbench",
"span_occurrence_ids": [
"V2-TERRA-03:S3:S3-A",
"V2-TERRA-03:S3:S3-B",
"V2-TERRA-03:S3:S3-C",
"V2-TERRA-03:S3:S3-D"
]
},
{
"citation_occurrence_id": "V2-TERRA-03:S4",
"correction_ids": [],
"edition_id": "edition:keplinger-vor-2025",
"license": "CC BY-NC 4.0",
"license_treatment": "quote-minimal attributed spans; no full-text redistribution",
"raw_source_id": "S4",
"raw_title": "Assessment of Deep Research for dermatology literature reviews: Deep concern over the hype",
"readback": {
"captured_bytes": 111993,
"captured_sha256": "abbaa9f22035230d13152a68aaeecd8e83e40867c9dae4acd1d547e36c866793",
"edition_id": "edition:keplinger-vor-2025",
"http_status": 200,
"media_type": "text/html; charset=utf-8",
"redirect_chain": [],
"requested_url": "https://pmc.ncbi.nlm.nih.gov/articles/PMC13109748/",
"resolved_url": "https://pmc.ncbi.nlm.nih.gov/articles/PMC13109748/",
"retrieval_status": "retrieved"
},
"requested_url": "https://pmc.ncbi.nlm.nih.gov/articles/PMC13109748/",
"resolution_status": "unresolved",
"run_id": "V2-TERRA-03",
"source_work_id": "work:keplinger-dermatology-audit",
"span_occurrence_ids": [
"V2-TERRA-03:S4:S4-A",
"V2-TERRA-03:S4:S4-B",
"V2-TERRA-03:S4:S4-C"
]
},
{
"citation_occurrence_id": "V2-TERRA-03:S6",
"correction_ids": [],
"edition_id": "edition:keplinger-vor-2025",
"license": "CC BY-NC 4.0",
"license_treatment": "quote-minimal attributed spans; no full-text redistribution",
"raw_source_id": "S6",
"raw_title": "PubMed record and Figure 1 caption for Assessment of Deep Research for dermatology literature reviews: Deep concern over the hype",
"readback": {
"captured_bytes": 5565,
"captured_sha256": "a46109544fe4ff4504fab5e97abea3cb7172367aba54a308937386394f0ff046",
"edition_id": "edition:keplinger-vor-2025",
"http_status": 203,
"media_type": "text/html; charset=utf-8",
"redirect_chain": [],
"requested_url": "https://pubmed.ncbi.nlm.nih.gov/40904191/",
"resolved_url": "https://pubmed.ncbi.nlm.nih.gov/40904191/",
"retrieval_status": "inaccessible"
},
"requested_url": "https://pubmed.ncbi.nlm.nih.gov/40904191/",
"resolution_status": "unresolved",
"run_id": "V2-TERRA-03",
"source_work_id": "work:keplinger-dermatology-audit",
"span_occurrence_ids": [
"V2-TERRA-03:S6:S6-A"
]
},
{
"citation_occurrence_id": "V2-TERRA-04:S1_deepresearchbench",
"correction_ids": [],
"edition_id": "edition:drbench-arxiv-v1",
"license": "CC BY 4.0",
"license_treatment": "quote-minimal attributed spans",
"raw_source_id": "S1_deepresearchbench",
"raw_title": "DeepResearch Bench: A Comprehensive Benchmark for Deep Research Agents",
"readback": {
"captured_bytes": 352002,
"captured_sha256": "9aa2894dbeaac30b23e7ffc8107a7f53b6e3855c8511838551de5bd7a422cc42",
"edition_id": "edition:drbench-arxiv-v1",
"http_status": 200,
"media_type": "text/html; charset=utf-8",
"redirect_chain": [],
"requested_url": "https://arxiv.org/html/2506.11763",
"resolved_url": "https://arxiv.org/html/2506.11763",
"retrieval_status": "retrieved"
},
"requested_url": "https://arxiv.org/html/2506.11763",
"resolution_status": "unresolved",
"run_id": "V2-TERRA-04",
"source_work_id": "work:deepresearch-bench-paper",
"span_occurrence_ids": [
"V2-TERRA-04:S1_deepresearchbench:S1_method",
"V2-TERRA-04:S1_deepresearchbench:S1_table1",
"V2-TERRA-04:S1_deepresearchbench:S1_metric",
"V2-TERRA-04:S1_deepresearchbench:S1_time"
]
},
{
"citation_occurrence_id": "V2-TERRA-04:S2_researcherbench",
"correction_ids": [],
"edition_id": "edition:researcherbench-arxiv-v1",
"license": "arXiv non-exclusive distribution license",
"license_treatment": "metadata, digest, and quote-minimal attributed spans only",
"raw_source_id": "S2_researcherbench",
"raw_title": "ResearcherBench: Evaluating Deep AI Research Systems on the Frontiers of Scientific Inquiry",
"readback": {
"captured_bytes": 280402,
"captured_sha256": "c080d7304274d70a39651f49297cc65910ca833d7ee6a3de3052370733e44ee3",
"edition_id": "edition:researcherbench-arxiv-v1",
"http_status": 200,
"media_type": "text/html; charset=utf-8",
"redirect_chain": [],
"requested_url": "https://arxiv.org/html/2507.16280",
"resolved_url": "https://arxiv.org/html/2507.16280",
"retrieval_status": "retrieved"
},
"requested_url": "https://arxiv.org/html/2507.16280",
"resolution_status": "unresolved",
"run_id": "V2-TERRA-04",
"source_work_id": "work:researcherbench",
"span_occurrence_ids": [
"V2-TERRA-04:S2_researcherbench:S2_method",
"V2-TERRA-04:S2_researcherbench:S2_table2",
"V2-TERRA-04:S2_researcherbench:S2_scope",
"V2-TERRA-04:S2_researcherbench:S2_time"
]
},
{
"citation_occurrence_id": "V2-TERRA-04:S3_reportbench",
"correction_ids": [],
"edition_id": "edition:reportbench-arxiv-v1",
"license": "CC BY 4.0",
"license_treatment": "quote-minimal attributed spans",
"raw_source_id": "S3_reportbench",
"raw_title": "ReportBench: Evaluating Deep Research Agents via Academic Survey Tasks",
"readback": {
"captured_bytes": 157386,
"captured_sha256": "055e568189a402dd510c7c84be60f59465d9a001808936bc25cb0026ac25d267",
"edition_id": "edition:reportbench-arxiv-v1",
"http_status": 200,
"media_type": "text/html; charset=utf-8",
"redirect_chain": [],
"requested_url": "https://arxiv.org/html/2508.15804",
"resolved_url": "https://arxiv.org/html/2508.15804",
"retrieval_status": "retrieved"
},
"requested_url": "https://arxiv.org/html/2508.15804",
"resolution_status": "unresolved",
"run_id": "V2-TERRA-04",
"source_work_id": "work:reportbench",
"span_occurrence_ids": [
"V2-TERRA-04:S3_reportbench:S3_method",
"V2-TERRA-04:S3_reportbench:S3_table1",
"V2-TERRA-04:S3_reportbench:S3_time",
"V2-TERRA-04:S3_reportbench:S3_limitations"
]
},
{
"citation_occurrence_id": "V2-TERRA-04:S4_urlhealth",
"correction_ids": [],
"edition_id": "edition:url-health-arxiv-v1",
"license": "CC0 1.0",
"license_treatment": "quote-minimal attributed spans",
"raw_source_id": "S4_urlhealth",
"raw_title": "Detecting and Correcting Reference Hallucinations in Commercial LLMs and Deep Research Agents",
"readback": {
"captured_bytes": 302498,
"captured_sha256": "332e5b5cb4b0ee7065b1bbc30436dfdfeafca8130fb87e32b0020026d8bbafe1",
"edition_id": "edition:url-health-arxiv-v1",
"http_status": 200,
"media_type": "text/html; charset=utf-8",
"redirect_chain": [],
"requested_url": "https://arxiv.org/html/2604.03173",
"resolved_url": "https://arxiv.org/html/2604.03173",
"retrieval_status": "retrieved"
},
"requested_url": "https://arxiv.org/html/2604.03173",
"resolution_status": "unresolved",
"run_id": "V2-TERRA-04",
"source_work_id": "work:url-health",
"span_occurrence_ids": [
"V2-TERRA-04:S4_urlhealth:S4_table2",
"V2-TERRA-04:S4_urlhealth:S4_comparison",
"V2-TERRA-04:S4_urlhealth:S4_method",
"V2-TERRA-04:S4_urlhealth:S4_limitations"
]
},
{
"citation_occurrence_id": "V2-TERRA-04:S5_cited_not_verified",
"correction_ids": [],
"edition_id": "edition:cnv-arxiv-v1",
"license": "CC BY 4.0",
"license_treatment": "quote-minimal attributed spans",
"raw_source_id": "S5_cited_not_verified",
"raw_title": "Cited but Not Verified: Parsing and Evaluating Source Attribution in LLM Deep Research Agents",
"readback": {
"captured_bytes": 138385,
"captured_sha256": "7c5e3c33f3122b07d176e6679cd6762a95a6babaf4a353686f019f905b5526ca",
"edition_id": "edition:cnv-arxiv-v1",
"http_status": 200,
"media_type": "text/html; charset=utf-8",
"redirect_chain": [],
"requested_url": "https://arxiv.org/html/2605.06635",
"resolved_url": "https://arxiv.org/html/2605.06635",
"retrieval_status": "retrieved"
},
"requested_url": "https://arxiv.org/html/2605.06635",
"resolution_status": "unresolved",
"run_id": "V2-TERRA-04",
"source_work_id": "work:cited-not-verified",
"span_occurrence_ids": [
"V2-TERRA-04:S5_cited_not_verified:S5_method",
"V2-TERRA-04:S5_cited_not_verified:S5_scope",
"V2-TERRA-04:S5_cited_not_verified:S5_table1",
"V2-TERRA-04:S5_cited_not_verified:S5_depth",
"V2-TERRA-04:S5_cited_not_verified:S5_limitations"
]
}
],
"unsupported_or_force_raised_occurrence_ids": [
"V2-SOL-01:r3_deepresearch_bench_fact",
"V2-SOL-01:r4_deeptrace_support",
"V2-SOL-02:R3_deepresearch_bench_fact",
"V2-SOL-02:R5_deeptrace_audit",
"V2-SOL-03:r3_deepresearch_bench_fact",
"V2-SOL-03:r4_deeptrace_support",
"V2-SOL-03:r8_search_depth_ablation",
"V2-SOL-03:r9_verifier_calibration",
"V2-SOL-04:R1",
"V2-SOL-04:R2",
"V2-SOL-04:R3",
"V2-SOL-04:R4",
"V2-SOL-04:R5",
"V2-SOL-04:R8",
"V2-TERRA-01:r1_deepresearchbench_fact",
"V2-TERRA-01:r4_cited_not_verified_source_attribution",
"V2-TERRA-02:answer",
"V2-TERRA-02:r1_deepresearchbench_fact",
"V2-TERRA-02:r4_deeptrace",
"V2-TERRA-03:R3"
]
},
"content_digest": "sha256:8400ea3bfe6c1e2516ab92710486eff3e62f5fee783a38859b383a561181533b",
"content_length": 132764,
"edition_label": "Deterministic projection of accepted EM-0026 relations",
"id": "em:dossier-edition:sha256:f0111e28ede7d9e17c435927cba483d0d9a7acf3dc23b38322ee4677d8cb70c0",
"key": "edition-em0026-audit-projection",
"media_type": "application/json",
"retrieved_at": "2026-08-23T18:33:06Z",
"visibility": "public",
"work_key": "work-em0026-audit-instrument"
}
],
"evaluations": [
{
"claim_family_key": "family-agent-citation-lineage",
"frontier": "research-candidate-em-0029",
"id": "em:dossier-evaluation:sha256:8065248caa97a4088854ab9c5d7b7729ad5e5e5d37234d757cd51e1dafb4503b",
"key": "evaluation-encyclopedia",
"label": "The bounded record contains empirical methods for URL resolution and claim-to-source support, but the eight reports reuse overlapping capture, source, method, and derivation lineages. Report agreement alone adds no independent warrant beyond the inspected source record.",
"policy_id": "em:application-policy:encyclopedia-v0.1",
"reason_codes": [
"citation-resolution-separated-from-support",
"shared-capture-lineage",
"candidate-warrants-scope-bounded",
"unresolved-citations-preserved"
],
"visibility": "public"
},
{
"claim_family_key": "family-agent-citation-lineage",
"frontier": "research-candidate-em-0029",
"id": "em:dossier-evaluation:sha256:17bf0049a751e753809793e2f89dfea413cda0c50049ce4329865241e2362091",
"key": "evaluation-skeptical",
"label": "This pilot cannot estimate current agent citation reliability: thirty-four citation occurrences remain unresolved, twenty claims required correction or no credit, four warrant groups remain pending, and zero warrant roots are independently confirmed by the packet. Inspect the exact source span before relying on a polished cited answer.",
"policy_id": "em:application-policy:skeptical-v0.1",
"reason_codes": [
"no-representative-agent-sample",
"unresolved-citation-set-material",
"semantic-strengthening-no-credit",
"warrant-independence-unconfirmed"
],
"visibility": "public"
}
],
"evidence_relations": [
{
"basis_span_keys": [
"span-audit-report-1",
"span-audit-capture-lineage"
],
"from_ref": "lineage-v2-sol-01-a73232acf8",
"id": "em:dossier-evidence-relation:sha256:abcc4d174a23391ada839ade500ca01bc54bba21adb6f16b792623606c50cb39",
"key": "relation-v2-sol-01-e7797662c8",
"note": "This report is a captured observation, not an independent evidence root.",
"relation_type": "dependence",
"to_ref": "lineage-capture-dependence-unknown",
"visibility": "public"
},
{
"basis_span_keys": [
"span-audit-report-2",
"span-audit-capture-lineage"
],
"from_ref": "lineage-v2-sol-02-82e3cd76cd",
"id": "em:dossier-evidence-relation:sha256:b4129699aedced2e4c8708199eef247fa1451d4a6189778a2e24d61ab0c373dc",
"key": "relation-v2-sol-02-8b059b3c99",
"note": "This report is a captured observation, not an independent evidence root.",
"relation_type": "dependence",
"to_ref": "lineage-capture-dependence-unknown",
"visibility": "public"
},
{
"basis_span_keys": [
"span-audit-report-3",
"span-audit-capture-lineage"
],
"from_ref": "lineage-v2-sol-03-a54afd37f7",
"id": "em:dossier-evidence-relation:sha256:1c2d0a17de3a97c42675d5061b130c73ac2100db232a5cb438800aa5349a228d",
"key": "relation-v2-sol-03-6b89558412",
"note": "This report is a captured observation, not an independent evidence root.",
"relation_type": "dependence",
"to_ref": "lineage-capture-dependence-unknown",
"visibility": "public"
},
{
"basis_span_keys": [
"span-audit-report-4",
"span-audit-capture-lineage"
],
"from_ref": "lineage-v2-sol-04-7c709ab690",
"id": "em:dossier-evidence-relation:sha256:8b5b57dc704e5af9085ae59c284443a864d80490fc3a1154d8112d48265f88dc",
"key": "relation-v2-sol-04-5e52e77bc3",
"note": "This report is a captured observation, not an independent evidence root.",
"relation_type": "dependence",
"to_ref": "lineage-capture-dependence-unknown",
"visibility": "public"
},
{
"basis_span_keys": [
"span-audit-report-5",
"span-audit-capture-lineage"
],
"from_ref": "lineage-v2-terra-01-962446cd82",
"id": "em:dossier-evidence-relation:sha256:07a3875663081289cb85d7862fa24a740a36e7268e8f10ec5e6ba78d299f2f08",
"key": "relation-v2-terra-01-d2701a39bc",
"note": "This report is a captured observation, not an independent evidence root.",
"relation_type": "dependence",
"to_ref": "lineage-capture-dependence-unknown",
"visibility": "public"
},
{
"basis_span_keys": [
"span-audit-report-6",
"span-audit-capture-lineage"
],
"from_ref": "lineage-v2-terra-02-7e359a7894",
"id": "em:dossier-evidence-relation:sha256:fbb42d9bedeae01728f8e2aa12153a785ecaf03680114da14242b27dca57504d",
"key": "relation-v2-terra-02-0ca185da71",
"note": "This report is a captured observation, not an independent evidence root.",
"relation_type": "dependence",
"to_ref": "lineage-capture-dependence-unknown",
"visibility": "public"
},
{
"basis_span_keys": [
"span-audit-report-7",
"span-audit-capture-lineage"
],
"from_ref": "lineage-v2-terra-03-c20170eb34",
"id": "em:dossier-evidence-relation:sha256:0a981ca96de2c12fe88ec3260e8b52822ca4b5236b6978492b564d57a859aa69",
"key": "relation-v2-terra-03-672f07dbef",
"note": "This report is a captured observation, not an independent evidence root.",
"relation_type": "dependence",
"to_ref": "lineage-capture-dependence-unknown",
"visibility": "public"
},
{
"basis_span_keys": [
"span-audit-report-8",
"span-audit-capture-lineage"
],
"from_ref": "lineage-v2-terra-04-d04d23915b",
"id": "em:dossier-evidence-relation:sha256:a6251888d77bee33da27dc27b2936c6aaff75b8d3783fcb2aeec8b407add2eea",
"key": "relation-v2-terra-04-b8e0fb7c26",
"note": "This report is a captured observation, not an independent evidence root.",
"relation_type": "dependence",
"to_ref": "lineage-capture-dependence-unknown",
"visibility": "public"
},
{
"basis_span_keys": [
"span-092745c7db6407d4d52927642741cde229fad05a5347bb18-7b04227f8f",
"span-39a9931f05ea356fe904ed15768fa08c66440b1fcb20f98e-191c24bbb2",
"span-46c392bc942cd88d525d74298e6bc6ebd9aeaec533ddfb7b-774f5b996e",
"span-8c8d9a6dc003bf0298542b85df3f6f08db631e68e89f606f-92731379f2",
"span-94ad583514aa7e04b8d5f7b8f9cdda6fdffe8ca6712f033d-63c1ba24eb",
"span-b38a2af5982b4c85bc205d2f533a23ed3a8f40a49b651a5f-27395a6291",
"span-b6a6ba48a2c4e352c8565b671a15a90400739f2aeefda8a6-e8232daf27",
"span-b903dc2284e1b4d80bb6ef948d79bbdbc7d7447e871281b8-6ef6234316",
"span-f05b3189dc3fafd45b5bde643a91fd0616832928215a37a2-47f72e77e4"
],
"from_ref": "assertion-keplinger-metadata-versus-support-73dfc141c5",
"id": "em:dossier-evidence-relation:sha256:cf9857ecb3c92201f72007257a23a62a68c8b655f551d3e62392baef87507e8a",
"key": "relation-keplinger-metadata-versus-support-c23e105c8b",
"note": "In one five-system dermatology-review audit, reference identifiability and metadata correctness did not establish sentence-level claim-citation concordance.",
"relation_type": "support",
"to_ref": "prop-keplinger-metadata-versus-support-73dfc141c5",
"visibility": "public"
},
{
"basis_span_keys": [
"span-audit-source-boundary"
],
"from_ref": "lineage-keplinger-metadata-versus-support-73dfc141c5",
"id": "em:dossier-evidence-relation:sha256:4b3a3d72252301adab5ab537f7a42b8deb03eef08dddb12a0a9d227287d89f9f",
"key": "relation-keplinger-metadata-versus-support-d9ee7c5a8a",
"note": "Candidate warrant count is not an independent-program count.",
"relation_type": "dependence",
"to_ref": "lineage-source-dependence-unknown",
"visibility": "public"
},
{
"basis_span_keys": [
"span-015bb68c904bf225e84c82b1acffb3cc38a4a087d7b0f79d-4c454092b1",
"span-0d4f357f2a5565bdaea01a1e537650958d71216868d1a2b0-9b0af35d76",
"span-13cb0ddb087b82988b7445a9cfb02562f9c044ef4f244e8b-87c3b42c68",
"span-196c54d6212fbc0c054dbf8e8467f2388c1f788bb93f4b1a-d460e927f6",
"span-279db725380c133044944dccb379b75a396242219aa30437-569fc9e2fd",
"span-2f311bfb451c6208101cb7583af37c338914fcdb10f5c75b-90e7764ce2",
"span-7d7331942f9d0520e18e96206e7718e5187788be4d48cef9-dfdbc1d019",
"span-92eb7cf19745e24190dda84c77d590941134d09f06805c0f-88afe43377",
"span-9fe42cb48702d8d96ebbbc0309692214ece6ec495c447c06-80da541506",
"span-b0a39c963cdfdc5f9852cfc9f7e33e05d31232bd1d899885-adf9545222",
"span-b83996e3b5fdf878e04d6d41d0e7a1eebee9fdd9bbccf2ae-820a3eab86",
"span-c27aaca02bda06ed764be53351158fc862af9d4a556d3e82-104e7cda52",
"span-e4a9f30effc031388ed5c9993dc6f071903bec5e3f1998db-5d95269b76",
"span-e5bcf0241a1aeae7d9a92755b1bcf0adeeea93a8922f22f8-77a5ff4cc0",
"span-eaf37bef7b8dc35f3e00b85d3e58ad872dc0ea604a062c8c-fa09300f85",
"span-ef637b38f36195bea09205d532e8744daf3e257f93d97144-26007673a3",
"span-f98f1274859ba98cc538e3be6d11d345a5382cfcb9516aa1-d60b765c48"
],
"from_ref": "assertion-deepresearch-bench-fact-ab2a1c0fd8",
"id": "em:dossier-evidence-relation:sha256:6adf0bc790adeb0f5b77cd93a3a1e1f7fbf3b4f2cd43d447246cd5c3a1d5afde",
"key": "relation-deepresearch-bench-fact-a3f61d7a3a",
"note": "DeepResearch Bench evaluated binary support for deduplicated statement-URL pairs; numeric values are edition-specific.",
"relation_type": "support",
"to_ref": "prop-deepresearch-bench-fact-ab2a1c0fd8",
"visibility": "public"
},
{
"basis_span_keys": [
"span-audit-source-boundary"
],
"from_ref": "lineage-deepresearch-bench-fact-ab2a1c0fd8",
"id": "em:dossier-evidence-relation:sha256:ac02cb9da6b4165d3964e987c72bb81fd0dd0716cc05338385baa1b16a3cc606",
"key": "relation-deepresearch-bench-fact-2efd170c89",
"note": "Candidate warrant count is not an independent-program count.",
"relation_type": "dependence",
"to_ref": "lineage-source-dependence-unknown",
"visibility": "public"
},
{
"basis_span_keys": [
"span-35927851edd2e7e5e0f60d98498940f88304ba99bf1f85a0-2fe586436b",
"span-8cce3d9ee1e1674ba1b93321dd9b99a55d834ca980451f8a-d2fceafe11",
"span-9cb1700c08368feb234207045354cbf559d6ed80f2b0bae8-f898722eb5",
"span-9dd6eb46529e77a62748093528e52685fbd92d46a756aef7-96d75d9e2b",
"span-a08468cb62f2feb21d9cbb95175ce00f54592bd83573cc91-03f65cfa45",
"span-a15ab02e9fcb23df02fb03d28967d4220ddf65ad380d2230-ddf85fa359",
"span-aa1534ce2e177153ab85e512deb8d064d14ec1eae16d8032-2299096a09",
"span-e2c9f6220843a04665f5d1cd142ac15f04211ace5c579fa6-c5e0476df0",
"span-supplement-deeptrace-gemini-prose-40-3-ed29752a6f",
"span-supplement-deeptrace-gemini-table-50-3-7eb7bd254b"
],
"from_ref": "assertion-deeptrace-support-variation-de1dfdaff3",
"id": "em:dossier-evidence-relation:sha256:6d44f25458001a3d2b303c223d05f31ec5d47f2a5870083593594141290a3a69",
"key": "relation-deeptrace-support-variation-1a84ef7c6f",
"note": "DeepTRACE measured statement-source support and found between-system variation; its Gemini value is internally inconsistent between table and prose.",
"relation_type": "support",
"to_ref": "prop-deeptrace-support-variation-de1dfdaff3",
"visibility": "public"
},
{
"basis_span_keys": [
"span-audit-source-boundary"
],
"from_ref": "lineage-deeptrace-support-variation-de1dfdaff3",
"id": "em:dossier-evidence-relation:sha256:d8b005710217434522fa0996a01ac3b30d46ab7680b55961c8abf04d798a3adf",
"key": "relation-deeptrace-support-variation-678e984a6a",
"note": "Candidate warrant count is not an independent-program count.",
"relation_type": "dependence",
"to_ref": "lineage-source-dependence-unknown",
"visibility": "public"
},
{
"basis_span_keys": [
"span-732c03e69c6e142a492275cee5d1e2d474c75422c9b84aaa-f7675998cc",
"span-74f6acafbb4c54d556a572460bf1b45bacd1e293f4aa16cc-5ee15c82c4",
"span-912c931063109f84ea88aea34891dd5e0f9147d2176af258-4d566ca076",
"span-bb16d5f06abe8641d9ea6694e549aa15e240dd88946aa47a-5946b6e6b7",
"span-cd21c450c0d33df20a2b1a539d5725347db1b5bab7d1e986-2e3e0c3ac7",
"span-deada8bacb987a47f2a42e52fbfeb5c1beb3d250c0be79e2-0e91f5cacc"
],
"from_ref": "assertion-cnv-link-relevance-support-gap-81b86b6d92",
"id": "em:dossier-evidence-relation:sha256:64e195a6456aebffbb478b57f1cefbe57b12973371ad65e8360feffcf08eae3d",
"key": "relation-cnv-link-relevance-support-gap-165443d270",
"note": "Cited but Not Verified operationalized link access, topical relevance, and factual support separately and reported materially different rates.",
"relation_type": "support",
"to_ref": "prop-cnv-link-relevance-support-gap-81b86b6d92",
"visibility": "public"
},
{
"basis_span_keys": [
"span-audit-source-boundary"
],
"from_ref": "lineage-cnv-link-relevance-support-gap-81b86b6d92",
"id": "em:dossier-evidence-relation:sha256:04b94120934511717d4eda8065675391ac59be7024d0594187572ff6909aff89",
"key": "relation-cnv-link-relevance-support-gap-911219619e",
"note": "Candidate warrant count is not an independent-program count.",
"relation_type": "dependence",
"to_ref": "lineage-source-dependence-unknown",
"visibility": "public"
},
{
"basis_span_keys": [
"span-417182b4c15ac550469e435a02aab1e91b395f72f186312d-e9e2c4a70c",
"span-6e961498ed0f81a950f775956587f741e0eb9f11640fd3cb-afa5b59854",
"span-c3dfc98566c27561eb7267972e00e3e1ba6a399a3ff5d002-37f40d3260",
"span-d638c8b21f0e4a0f840880b6ffe0217ce1623758cc09fdcc-8585f6d39b"
],
"from_ref": "assertion-cnv-search-depth-ablation-dad05fa3f9",
"id": "em:dossier-evidence-relation:sha256:454a592bc79732fb100cc628b9b5c7caa7eef3358f0a3acefc85d7e55147fdba",
"key": "relation-cnv-search-depth-ablation-9c01321092",
"note": "Within the Cited but Not Verified harness, the 2-to-150-call ablation reduced reported Fact Check scores for two setups without establishing a general causal law about search depth.",
"relation_type": "support",
"to_ref": "prop-cnv-search-depth-ablation-dad05fa3f9",
"visibility": "public"
},
{
"basis_span_keys": [
"span-audit-source-boundary"
],
"from_ref": "lineage-cnv-search-depth-ablation-dad05fa3f9",
"id": "em:dossier-evidence-relation:sha256:7926587a91336078c7113565325881a8ab16d9fa82b4aae27b1c662c7e87f421",
"key": "relation-cnv-search-depth-ablation-d7aecf2d25",
"note": "Candidate warrant count is not an independent-program count.",
"relation_type": "dependence",
"to_ref": "lineage-source-dependence-unknown",
"visibility": "public"
},
{
"basis_span_keys": [
"span-55de6a6ca9414a252592e9f77544a9259533dd713ce1f12f-0826216ffc",
"span-a18e778fac3d49e81167f05e09fbc361e91af8c0c01b0fec-ca1b18e206",
"span-bb62928b5e0cb4e373b2f6bfeb579d8564d05da5588f8b15-771310f4ef",
"span-supplement-url-health-drbench-precollected-outputs-44923dff13"
],
"from_ref": "assertion-url-health-resolution-9fcccc534a",
"id": "em:dossier-evidence-relation:sha256:fd5741fb1196a75091a20a03b6e903c1fdc7e338c94d9f366d8cec6caf3613fe",
"key": "relation-url-health-resolution-0b26fb45c7",
"note": "The URL-health study measured HTTP resolution and Wayback presence, not semantic claim support, on outputs including reused DeepResearch Bench material.",
"relation_type": "support",
"to_ref": "prop-url-health-resolution-9fcccc534a",
"visibility": "public"
},
{
"basis_span_keys": [
"span-audit-source-boundary"
],
"from_ref": "lineage-url-health-resolution-9fcccc534a",
"id": "em:dossier-evidence-relation:sha256:a031938c655ded3d1f19e577057875acc6f3c243190453440c6582eefc8473d8",
"key": "relation-url-health-resolution-b1cc744437",
"note": "Candidate warrant count is not an independent-program count.",
"relation_type": "dependence",
"to_ref": "lineage-source-dependence-unknown",
"visibility": "public"
},
{
"basis_span_keys": [
"span-1db12b68cfeb4c5f5a96fa9cd2e2bd46e8dfb4f096f73154-cea47f0c65",
"span-3015a109e7898cb720e426e9a57c526277b648c75f1fb4b1-745d84ff44",
"span-6c1d327e58ef001533ea1c11010e974436bbeafbde138efc-f6efd8d376",
"span-7a6046a756d6ec0315d330cf0817d062ea48498bdc9c90b3-c3a1e80029",
"span-9c7a4c186cc85f185aa293a7c5a46c08f9dbbc1a9e63c449-feff07a6e2",
"span-ba6b8a8b7d9121ebb05d594176ceed3d1e9dbc637876783a-d85147b6b0",
"span-ddf4320a6d302c13d28b6225eb7e79ac15583349e1b449d4-f89bb46179",
"span-e223640cb0ed12c87ca1d2406f3276f30a3b8d2017dd4dc1-3cfecd76bf"
],
"from_ref": "assertion-reportbench-match-rate-46ddf0a876",
"id": "em:dossier-evidence-relation:sha256:ce23bc99cef3ca2b0b0e3597ada68536bf0813903e3468c894914c79a30c6d41",
"key": "relation-reportbench-match-rate-6964e41a69",
"note": "ReportBench compared cited statements with retrieved cited-page content and reported sub-100-percent semantic match rates in its bounded survey-task setting.",
"relation_type": "support",
"to_ref": "prop-reportbench-match-rate-46ddf0a876",
"visibility": "public"
},
{
"basis_span_keys": [
"span-audit-source-boundary"
],
"from_ref": "lineage-reportbench-match-rate-46ddf0a876",
"id": "em:dossier-evidence-relation:sha256:e0d55a7407bab1cd8408d1d0d7fdbfd4858f288a864baa9d7442664633792215",
"key": "relation-reportbench-match-rate-4f30573b54",
"note": "Candidate warrant count is not an independent-program count.",
"relation_type": "dependence",
"to_ref": "lineage-source-dependence-unknown",
"visibility": "public"
},
{
"basis_span_keys": [
"span-audit-pending-1"
],
"from_ref": "assertion-citation-verifier-calibration-5b40082faa",
"id": "em:dossier-evidence-relation:sha256:ef8f517c3d030147f56936e279e6260a63f142b3514a5e98475bb83e95940d21",
"key": "relation-citation-verifier-calibration-5b40082faa",
"note": "The captured spans do not semantically close the normalized proposition; the warrant remains pending and receives no credit.",
"relation_type": "qualification",
"to_ref": "prop-citation-verifier-calibration-da47a19a65",
"visibility": "public"
},
{
"basis_span_keys": [
"span-audit-pending-2"
],
"from_ref": "assertion-liveresearchbench-e1-e2-e3-5fb5d671f7",
"id": "em:dossier-evidence-relation:sha256:d8f8705949761213814c9a9c1c47dca7d298d10a0c2f763c5c6ac6c6a372c880",
"key": "relation-liveresearchbench-e1-e2-e3-5fb5d671f7",
"note": "The captured spans do not semantically close the normalized proposition; the warrant remains pending and receives no credit.",
"relation_type": "qualification",
"to_ref": "prop-liveresearchbench-e1-e2-e3-07e84c4837",
"visibility": "public"
},
{
"basis_span_keys": [
"span-audit-pending-3"
],
"from_ref": "assertion-researcherbench-faithfulness-groundedness-c84dcaee54",
"id": "em:dossier-evidence-relation:sha256:c059f728a9e941f48d872e2b9acd4d1c7019a7428045412234b4aa6f44a4a30f",
"key": "relation-researcherbench-faithfulness-groundedness-c84dcaee54",
"note": "The captured spans do not semantically close the normalized proposition; the warrant remains pending and receives no credit.",
"relation_type": "qualification",
"to_ref": "prop-researcherbench-faithfulness-groundedness-0deb7adf49",
"visibility": "public"
},
{
"basis_span_keys": [
"span-audit-pending-4"
],
"from_ref": "assertion-url-health-correction-loop-eb8a6384c6",
"id": "em:dossier-evidence-relation:sha256:22dd32ad9045555bbab3a8c52380e5dd71c6d6769f24ce61436bbcdbc582a736",
"key": "relation-url-health-correction-loop-eb8a6384c6",
"note": "The captured spans do not semantically close the normalized proposition; the warrant remains pending and receives no credit.",
"relation_type": "qualification",
"to_ref": "prop-url-health-correction-loop-f63084450b",
"visibility": "public"
},
{
"basis_span_keys": [
"span-audit-counts"
],
"from_ref": "assertion-derived-counts",
"id": "em:dossier-evidence-relation:sha256:20a26ae3bab2bc8e4b117f132d539dc7579243b441eb81d8321e39f30b0f8ab9",
"key": "relation-assertion-derived-counts-234c8dbdaf",
"note": "Each displayed total is derived from typed packet records under the stored grammar.",
"relation_type": "support",
"to_ref": "prop-relation-derived-counts",
"visibility": "public"
},
{
"basis_span_keys": [
"span-audit-unresolved"
],
"from_ref": "assertion-unresolved-citations",
"id": "em:dossier-evidence-relation:sha256:ec1bf32704aec42df6058d83a4f445cdd437060ee24b66f898d73b6453556b12",
"key": "relation-assertion-unresolved-citations-35a17beebf",
"note": "The complete unresolved set remains visible and receives no warrant credit.",
"relation_type": "undercutting",
"to_ref": "prop-agreement-not-independent-warrant",
"visibility": "public"
},
{
"basis_span_keys": [
"span-audit-inaccessible"
],
"from_ref": "assertion-inaccessible-carriers",
"id": "em:dossier-evidence-relation:sha256:ee327e39649c9b98b149d1e3e92d96dda939305577d8778b917864488dea59c7",
"key": "relation-assertion-inaccessible-carriers-a9d4494367",
"note": "The three inaccessible carrier occurrences remain visible and receive no credit.",
"relation_type": "undercutting",
"to_ref": "prop-agreement-not-independent-warrant",
"visibility": "public"
},
{
"basis_span_keys": [
"span-audit-unsupported"
],
"from_ref": "assertion-unsupported-claims",
"id": "em:dossier-evidence-relation:sha256:19686631a3f82281887dfbd358b53954fa3b089b46b2723e519ddb64238ab382",
"key": "relation-assertion-unsupported-claims-30720e36be",
"note": "The complete unsupported or force-raised set remains visible and receives no credit.",
"relation_type": "undercutting",
"to_ref": "prop-agreement-not-independent-warrant",
"visibility": "public"
},
{
"basis_span_keys": [
"span-audit-rejected"
],
"from_ref": "assertion-independently-rejected-claims",
"id": "em:dossier-evidence-relation:sha256:d242a9ec43c1658f686946adc4afcd0377826066d64dd6b7acf9edc10ed507d6",
"key": "relation-assertion-independently-rejected-claims-4af09a9196",
"note": "All nine independently rejected claims remain visible and receive no credit.",
"relation_type": "undercutting",
"to_ref": "prop-agreement-not-independent-warrant",
"visibility": "public"
}
],
"format": "epistemedia-dossier-v0.1",
"lineages": [
{
"assertion_keys": [],
"basis_span_keys": [
"span-audit-capture-lineage"
],
"depends_on": [],
"dimensions": [
"model",
"retrieval",
"prompt",
"other"
],
"id": "em:dossier-lineage:sha256:3815b6a71162f7b71a985c44247317779c82180778eaaa708a748c5bc867fcf5",
"key": "lineage-capture-dependence-unknown",
"note": "Unknown provider and retrieval dependencies remain; all reports share the exact prompt and one bounded capture program, so run multiplicity gets zero automatic independence credit.",
"status": "unknown",
"visibility": "public"
},
{
"assertion_keys": [],
"basis_span_keys": [
"span-audit-source-boundary"
],
"depends_on": [],
"dimensions": [
"source",
"data",
"method",
"retrieval",
"other"
],
"id": "em:dossier-lineage:sha256:b6387bfc3857ba2a508baea6b6fa50d557c844ea09d11d231b862921cd0da13d",
"key": "lineage-source-dependence-unknown",
"note": "Unknown residual independence remains across task data, judge methods, retrieval, source, edition, span, derivation, and upstream-citation lineages.",
"status": "unknown",
"visibility": "public"
},
{
"assertion_keys": [
"assertion-v2-sol-01-a73232acf8"
],
"basis_span_keys": [
"span-audit-report-1",
"span-audit-capture-lineage"
],
"depends_on": [
"lineage-capture-dependence-unknown"
],
"dimensions": [
"model",
"retrieval",
"prompt"
],
"id": "em:dossier-lineage:sha256:eaae15e292d35333b3ab5f2629d26645b56eb974d9b94c4be66394ea98adfc6b",
"key": "lineage-v2-sol-01-a73232acf8",
"note": "V2-SOL-01 shares the frozen prompt and capture program; requested profile does not establish an independent observer.",
"status": "known",
"visibility": "public"
},
{
"assertion_keys": [
"assertion-v2-sol-02-82e3cd76cd"
],
"basis_span_keys": [
"span-audit-report-2",
"span-audit-capture-lineage"
],
"depends_on": [
"lineage-capture-dependence-unknown"
],
"dimensions": [
"model",
"retrieval",
"prompt"
],
"id": "em:dossier-lineage:sha256:375018e7e650004d75ee6ec790a307e920b5b99550ea561029f890d077a78de4",
"key": "lineage-v2-sol-02-82e3cd76cd",
"note": "V2-SOL-02 shares the frozen prompt and capture program; requested profile does not establish an independent observer.",
"status": "known",
"visibility": "public"
},
{
"assertion_keys": [
"assertion-v2-sol-03-a54afd37f7"
],
"basis_span_keys": [
"span-audit-report-3",
"span-audit-capture-lineage"
],
"depends_on": [
"lineage-capture-dependence-unknown"
],
"dimensions": [
"model",
"retrieval",
"prompt"
],
"id": "em:dossier-lineage:sha256:a9bd8d77d472a23659971cbeede225a04cee441b5421649a88dc49386ba2e306",
"key": "lineage-v2-sol-03-a54afd37f7",
"note": "V2-SOL-03 shares the frozen prompt and capture program; requested profile does not establish an independent observer.",
"status": "known",
"visibility": "public"
},
{
"assertion_keys": [
"assertion-v2-sol-04-7c709ab690"
],
"basis_span_keys": [
"span-audit-report-4",
"span-audit-capture-lineage"
],
"depends_on": [
"lineage-capture-dependence-unknown"
],
"dimensions": [
"model",
"retrieval",
"prompt"
],
"id": "em:dossier-lineage:sha256:0c40c79529de0273d8f444688a011b654bd77ccc266117a5555efdf61cc56f9b",
"key": "lineage-v2-sol-04-7c709ab690",
"note": "V2-SOL-04 shares the frozen prompt and capture program; requested profile does not establish an independent observer.",
"status": "known",
"visibility": "public"
},
{
"assertion_keys": [
"assertion-v2-terra-01-962446cd82"
],
"basis_span_keys": [
"span-audit-report-5",
"span-audit-capture-lineage"
],
"depends_on": [
"lineage-capture-dependence-unknown"
],
"dimensions": [
"model",
"retrieval",
"prompt"
],
"id": "em:dossier-lineage:sha256:44ab752637a55d129eea73b7dc64af33f6f02ff4d53133fbcfa69c5ccabfe297",
"key": "lineage-v2-terra-01-962446cd82",
"note": "V2-TERRA-01 shares the frozen prompt and capture program; requested profile does not establish an independent observer.",
"status": "known",
"visibility": "public"
},
{
"assertion_keys": [
"assertion-v2-terra-02-7e359a7894"
],
"basis_span_keys": [
"span-audit-report-6",
"span-audit-capture-lineage"
],
"depends_on": [
"lineage-capture-dependence-unknown"
],
"dimensions": [
"model",
"retrieval",
"prompt"
],
"id": "em:dossier-lineage:sha256:94588a3e1d333b8cd3934e12553510402161bcb9732b5b4b6f0fde471d5e8660",
"key": "lineage-v2-terra-02-7e359a7894",
"note": "V2-TERRA-02 shares the frozen prompt and capture program; requested profile does not establish an independent observer.",
"status": "known",
"visibility": "public"
},
{
"assertion_keys": [
"assertion-v2-terra-03-c20170eb34"
],
"basis_span_keys": [
"span-audit-report-7",
"span-audit-capture-lineage"
],
"depends_on": [
"lineage-capture-dependence-unknown"
],
"dimensions": [
"model",
"retrieval",
"prompt"
],
"id": "em:dossier-lineage:sha256:b136e05c65ba23d81c47c490f981f5d14188fe336874a3451d53d4304399d210",
"key": "lineage-v2-terra-03-c20170eb34",
"note": "V2-TERRA-03 shares the frozen prompt and capture program; requested profile does not establish an independent observer.",
"status": "known",
"visibility": "public"
},
{
"assertion_keys": [
"assertion-v2-terra-04-d04d23915b"
],
"basis_span_keys": [
"span-audit-report-8",
"span-audit-capture-lineage"
],
"depends_on": [
"lineage-capture-dependence-unknown"
],
"dimensions": [
"model",
"retrieval",
"prompt"
],
"id": "em:dossier-lineage:sha256:a50f158019a6ee7aff96bda0fe111debe9b9ee294acc4e59fed858078ab52943",
"key": "lineage-v2-terra-04-d04d23915b",
"note": "V2-TERRA-04 shares the frozen prompt and capture program; requested profile does not establish an independent observer.",
"status": "known",
"visibility": "public"
},
{
"assertion_keys": [
"assertion-keplinger-metadata-versus-support-73dfc141c5"
],
"basis_span_keys": [
"span-092745c7db6407d4d52927642741cde229fad05a5347bb18-7b04227f8f",
"span-39a9931f05ea356fe904ed15768fa08c66440b1fcb20f98e-191c24bbb2",
"span-46c392bc942cd88d525d74298e6bc6ebd9aeaec533ddfb7b-774f5b996e",
"span-8c8d9a6dc003bf0298542b85df3f6f08db631e68e89f606f-92731379f2",
"span-94ad583514aa7e04b8d5f7b8f9cdda6fdffe8ca6712f033d-63c1ba24eb",
"span-b38a2af5982b4c85bc205d2f533a23ed3a8f40a49b651a5f-27395a6291",
"span-b6a6ba48a2c4e352c8565b671a15a90400739f2aeefda8a6-e8232daf27",
"span-b903dc2284e1b4d80bb6ef948d79bbdbc7d7447e871281b8-6ef6234316",
"span-f05b3189dc3fafd45b5bde643a91fd0616832928215a37a2-47f72e77e4",
"span-audit-source-boundary"
],
"depends_on": [
"lineage-source-dependence-unknown"
],
"dimensions": [
"source",
"data",
"method",
"retrieval",
"other"
],
"id": "em:dossier-lineage:sha256:a46612a0397294b1e58fa90ce5cfacd98ff271164d2a49eb549ee40b26b23225",
"key": "lineage-keplinger-metadata-versus-support-73dfc141c5",
"note": "Candidate only: data:keplinger-three-run-five-system-audit; method:human-reference-and-sentence-concordance-audit; derivation:keplinger-published-tables-and-figure. Residual independence is unknown.",
"status": "known",
"visibility": "public"
},
{
"assertion_keys": [
"assertion-deepresearch-bench-fact-ab2a1c0fd8"
],
"basis_span_keys": [
"span-015bb68c904bf225e84c82b1acffb3cc38a4a087d7b0f79d-4c454092b1",
"span-0d4f357f2a5565bdaea01a1e537650958d71216868d1a2b0-9b0af35d76",
"span-13cb0ddb087b82988b7445a9cfb02562f9c044ef4f244e8b-87c3b42c68",
"span-196c54d6212fbc0c054dbf8e8467f2388c1f788bb93f4b1a-d460e927f6",
"span-279db725380c133044944dccb379b75a396242219aa30437-569fc9e2fd",
"span-2f311bfb451c6208101cb7583af37c338914fcdb10f5c75b-90e7764ce2",
"span-7d7331942f9d0520e18e96206e7718e5187788be4d48cef9-dfdbc1d019",
"span-92eb7cf19745e24190dda84c77d590941134d09f06805c0f-88afe43377",
"span-9fe42cb48702d8d96ebbbc0309692214ece6ec495c447c06-80da541506",
"span-b0a39c963cdfdc5f9852cfc9f7e33e05d31232bd1d899885-adf9545222",
"span-b83996e3b5fdf878e04d6d41d0e7a1eebee9fdd9bbccf2ae-820a3eab86",
"span-c27aaca02bda06ed764be53351158fc862af9d4a556d3e82-104e7cda52",
"span-e4a9f30effc031388ed5c9993dc6f071903bec5e3f1998db-5d95269b76",
"span-e5bcf0241a1aeae7d9a92755b1bcf0adeeea93a8922f22f8-77a5ff4cc0",
"span-eaf37bef7b8dc35f3e00b85d3e58ad872dc0ea604a062c8c-fa09300f85",
"span-ef637b38f36195bea09205d532e8744daf3e257f93d97144-26007673a3",
"span-f98f1274859ba98cc538e3be6d11d345a5382cfcb9516aa1-d60b765c48",
"span-audit-source-boundary"
],
"depends_on": [
"lineage-source-dependence-unknown"
],
"dimensions": [
"source",
"data",
"method",
"retrieval",
"other"
],
"id": "em:dossier-lineage:sha256:e5dca8785a2cf9d98885a01cf163bc158cabd7a6b12008eccc6891009bf32391",
"key": "lineage-deepresearch-bench-fact-ab2a1c0fd8",
"note": "Candidate only: data:deepresearch-bench-100-tasks-and-system-outputs; method:fact-statement-url-support-evaluator; derivation:deepresearch-bench-edition-specific-tables. Residual independence is unknown.",
"status": "known",
"visibility": "public"
},
{
"assertion_keys": [
"assertion-deeptrace-support-variation-de1dfdaff3"
],
"basis_span_keys": [
"span-35927851edd2e7e5e0f60d98498940f88304ba99bf1f85a0-2fe586436b",
"span-8cce3d9ee1e1674ba1b93321dd9b99a55d834ca980451f8a-d2fceafe11",
"span-9cb1700c08368feb234207045354cbf559d6ed80f2b0bae8-f898722eb5",
"span-9dd6eb46529e77a62748093528e52685fbd92d46a756aef7-96d75d9e2b",
"span-a08468cb62f2feb21d9cbb95175ce00f54592bd83573cc91-03f65cfa45",
"span-a15ab02e9fcb23df02fb03d28967d4220ddf65ad380d2230-ddf85fa359",
"span-aa1534ce2e177153ab85e512deb8d064d14ec1eae16d8032-2299096a09",
"span-e2c9f6220843a04665f5d1cd142ac15f04211ace5c579fa6-c5e0476df0",
"span-supplement-deeptrace-gemini-prose-40-3-ed29752a6f",
"span-supplement-deeptrace-gemini-table-50-3-7eb7bd254b",
"span-audit-source-boundary"
],
"depends_on": [
"lineage-source-dependence-unknown"
],
"dimensions": [
"source",
"data",
"method",
"retrieval",
"other"
],
"id": "em:dossier-lineage:sha256:76899e048e8624bdbc295f77eba9d647f354f8bc235d5052ed62e4356ae20a90",
"key": "lineage-deeptrace-support-variation-de1dfdaff3",
"note": "Candidate only: data:deeptrace-303-questions-2727-outputs; method:deeptrace-statement-source-matrix; derivation:deeptrace-tables-and-prose. Residual independence is unknown.",
"status": "known",
"visibility": "public"
},
{
"assertion_keys": [
"assertion-cnv-link-relevance-support-gap-81b86b6d92"
],
"basis_span_keys": [
"span-732c03e69c6e142a492275cee5d1e2d474c75422c9b84aaa-f7675998cc",
"span-74f6acafbb4c54d556a572460bf1b45bacd1e293f4aa16cc-5ee15c82c4",
"span-912c931063109f84ea88aea34891dd5e0f9147d2176af258-4d566ca076",
"span-bb16d5f06abe8641d9ea6694e549aa15e240dd88946aa47a-5946b6e6b7",
"span-cd21c450c0d33df20a2b1a539d5725347db1b5bab7d1e986-2e3e0c3ac7",
"span-deada8bacb987a47f2a42e52fbfeb5c1beb3d250c0be79e2-0e91f5cacc",
"span-audit-source-boundary"
],
"depends_on": [
"lineage-source-dependence-unknown"
],
"dimensions": [
"source",
"data",
"method",
"retrieval",
"other"
],
"id": "em:dossier-lineage:sha256:ca155e4894daf1886d38f9d2e13d1757164295ffc0315af008cef7f895d71cbf",
"key": "lineage-cnv-link-relevance-support-gap-81b86b6d92",
"note": "Candidate only: data:cnv-130-drbench-and-browsecomp-queries; method:cnv-ast-access-relevance-fact-check; derivation:cnv-main-table. Residual independence is unknown.",
"status": "known",
"visibility": "public"
},
{
"assertion_keys": [
"assertion-cnv-search-depth-ablation-dad05fa3f9"
],
"basis_span_keys": [
"span-417182b4c15ac550469e435a02aab1e91b395f72f186312d-e9e2c4a70c",
"span-6e961498ed0f81a950f775956587f741e0eb9f11640fd3cb-afa5b59854",
"span-c3dfc98566c27561eb7267972e00e3e1ba6a399a3ff5d002-37f40d3260",
"span-d638c8b21f0e4a0f840880b6ffe0217ce1623758cc09fdcc-8585f6d39b",
"span-audit-source-boundary"
],
"depends_on": [
"lineage-source-dependence-unknown"
],
"dimensions": [
"source",
"data",
"method",
"retrieval",
"other"
],
"id": "em:dossier-lineage:sha256:959881cc29e06594b7814995687514cd49a424d23eed0c27263a3c34b6a67cff",
"key": "lineage-cnv-search-depth-ablation-dad05fa3f9",
"note": "Candidate only: data:cnv-depth-ablation-two-model-setups; method:cnv-ast-access-relevance-fact-check; derivation:cnv-depth-ablation-tables. Residual independence is unknown.",
"status": "known",
"visibility": "public"
},
{
"assertion_keys": [
"assertion-url-health-resolution-9fcccc534a"
],
"basis_span_keys": [
"span-55de6a6ca9414a252592e9f77544a9259533dd713ce1f12f-0826216ffc",
"span-a18e778fac3d49e81167f05e09fbc361e91af8c0c01b0fec-ca1b18e206",
"span-bb62928b5e0cb4e373b2f6bfeb579d8564d05da5588f8b15-771310f4ef",
"span-supplement-url-health-drbench-precollected-outputs-44923dff13",
"span-audit-source-boundary"
],
"depends_on": [
"lineage-source-dependence-unknown"
],
"dimensions": [
"source",
"data",
"method",
"retrieval",
"other"
],
"id": "em:dossier-lineage:sha256:92d2ecbee8e25a15eeb7a6c7fa6a54f79536633180a2740eda32d3b43027e580",
"key": "lineage-url-health-resolution-9fcccc534a",
"note": "Candidate only: data:url-health-drbench-and-expertqa-urls; method:http-resolution-wayback-classification; derivation:url-health-observational-results. Residual independence is unknown.",
"status": "known",
"visibility": "public"
},
{
"assertion_keys": [
"assertion-reportbench-match-rate-46ddf0a876"
],
"basis_span_keys": [
"span-1db12b68cfeb4c5f5a96fa9cd2e2bd46e8dfb4f096f73154-cea47f0c65",
"span-3015a109e7898cb720e426e9a57c526277b648c75f1fb4b1-745d84ff44",
"span-6c1d327e58ef001533ea1c11010e974436bbeafbde138efc-f6efd8d376",
"span-7a6046a756d6ec0315d330cf0817d062ea48498bdc9c90b3-c3a1e80029",
"span-9c7a4c186cc85f185aa293a7c5a46c08f9dbbc1a9e63c449-feff07a6e2",
"span-ba6b8a8b7d9121ebb05d594176ceed3d1e9dbc637876783a-d85147b6b0",
"span-ddf4320a6d302c13d28b6225eb7e79ac15583349e1b449d4-f89bb46179",
"span-e223640cb0ed12c87ca1d2406f3276f30a3b8d2017dd4dc1-3cfecd76bf",
"span-audit-source-boundary"
],
"depends_on": [
"lineage-source-dependence-unknown"
],
"dimensions": [
"source",
"data",
"method",
"retrieval",
"other"
],
"id": "em:dossier-lineage:sha256:a7a4824a78a41b9491f399ac0fe498624111accf239f3d2cac5bd00c3fb8813c",
"key": "lineage-reportbench-match-rate-46ddf0a876",
"note": "Candidate only: data:reportbench-100-survey-tasks; method:reportbench-statement-source-semantic-match; derivation:reportbench-table-1. Residual independence is unknown.",
"status": "known",
"visibility": "public"
},
{
"assertion_keys": [
"assertion-citation-verifier-calibration-5b40082faa"
],
"basis_span_keys": [
"span-audit-pending-1"
],
"depends_on": [],
"dimensions": [
"source",
"method",
"other"
],
"id": "em:dossier-lineage:sha256:3ecc0fb40cb67b7257029075c5c38aa9b62cc0d54b56e80661d3aec6987ce2d9",
"key": "lineage-citation-verifier-calibration-5b40082faa",
"note": "Unknown warrant closure: pending-insufficient-semantic-span-closure.",
"status": "unknown",
"visibility": "public"
},
{
"assertion_keys": [
"assertion-liveresearchbench-e1-e2-e3-5fb5d671f7"
],
"basis_span_keys": [
"span-audit-pending-2"
],
"depends_on": [],
"dimensions": [
"source",
"method",
"other"
],
"id": "em:dossier-lineage:sha256:225a03fc4bc72c1d554e2ebf41cc5cf4be1667dace9793dba2f166bfaedd9700",
"key": "lineage-liveresearchbench-e1-e2-e3-5fb5d671f7",
"note": "Unknown warrant closure: pending-insufficient-semantic-span-closure.",
"status": "unknown",
"visibility": "public"
},
{
"assertion_keys": [
"assertion-researcherbench-faithfulness-groundedness-c84dcaee54"
],
"basis_span_keys": [
"span-audit-pending-3"
],
"depends_on": [],
"dimensions": [
"source",
"method",
"other"
],
"id": "em:dossier-lineage:sha256:c44f7302ad6fa0a76bd58f42d85d669a27ea5c0ae442505d29605cc235c42d71",
"key": "lineage-researcherbench-faithfulness-groundedness-c84dcaee54",
"note": "Unknown warrant closure: pending-insufficient-semantic-span-closure.",
"status": "unknown",
"visibility": "public"
},
{
"assertion_keys": [
"assertion-url-health-correction-loop-eb8a6384c6"
],
"basis_span_keys": [
"span-audit-pending-4"
],
"depends_on": [],
"dimensions": [
"source",
"method",
"other"
],
"id": "em:dossier-lineage:sha256:1ab988a3e616af652fad84f23210936d86a59eaeb97a005e3d2d6582d6d89290",
"key": "lineage-url-health-correction-loop-eb8a6384c6",
"note": "Unknown warrant closure: pending-insufficient-semantic-span-closure.",
"status": "unknown",
"visibility": "public"
},
{
"assertion_keys": [
"assertion-derived-counts",
"assertion-unresolved-citations",
"assertion-inaccessible-carriers",
"assertion-unsupported-claims",
"assertion-independently-rejected-claims"
],
"basis_span_keys": [
"span-audit-counts",
"span-audit-unresolved",
"span-audit-inaccessible",
"span-audit-unsupported",
"span-audit-rejected"
],
"depends_on": [],
"dimensions": [
"apparatus",
"other"
],
"id": "em:dossier-lineage:sha256:8681d3455a6f3195212409272585e54585ec0a4858cc2d7d5e80535d2752a7df",
"key": "lineage-em0026-audit",
"note": "Deterministic relation audit over the accepted EM-0026 packet.",
"status": "known",
"visibility": "public"
}
],
"propositions": [
{
"id": "em:dossier-proposition:sha256:67a7bd889acc6ad90dfd59b814ba8ac1762359b06528a6b7d2ee2375b2ac6640",
"key": "prop-agreement-not-independent-warrant",
"scope": "The eight public-by-design EM-0026 captures and evidence through 2026-08-22.",
"text": "Separate agent reports do not by themselves supply independent evidentiary warrant when their prompt, runtime, retrieval, source, method, or derivation lineages overlap or remain unknown.",
"visibility": "public"
},
{
"id": "em:dossier-proposition:sha256:9ea4b2fe08fe6cfaeba34e4a5e5c20675e294739f746c601414da49a84e5f081",
"key": "prop-report-citation-observation",
"scope": "One captured run artifact; no claim of independence or generality.",
"text": "One frozen EM-0026 run produced a terminal citation-bearing report.",
"visibility": "public"
},
{
"id": "em:dossier-proposition:sha256:96e76b77a40ab697aba1a707480155b4a3285ff3444aae93ea691d81e5c7f246",
"key": "prop-unresolved-citations-no-credit",
"scope": "The accepted EM-0026 citation-occurrence ledger.",
"text": "Citation occurrences with inaccessible carriers, unmatched spans, or unresolved identity or warrant corrections receive no evidentiary credit.",
"visibility": "public"
},
{
"id": "em:dossier-proposition:sha256:c3d50556a71c22fdaaa8a9cf5c33b49d8b57c3b7fc01af73e23f30554ce23f61",
"key": "prop-unsupported-claims-no-credit",
"scope": "The accepted EM-0026 correction and semantic-review records.",
"text": "Agent claim occurrences strengthened beyond their linked source semantics remain visible and receive no warrant credit for the stronger wording.",
"visibility": "public"
},
{
"id": "em:dossier-proposition:sha256:3bf96c8701bbe2577c95293cef4b947e8596513cbe6d97f7c51bdc8255495808",
"key": "prop-inaccessible-carriers-no-credit",
"scope": "The accepted EM-0026 citation-occurrence ledger.",
"text": "Three cited carriers were inaccessible during accepted independent readback and receive no carrier-level credit even where another edition of the work exists.",
"visibility": "public"
},
{
"id": "em:dossier-proposition:sha256:b81af4e7609aa37357f85a06a2d95a9e5c25f159a67b91694a13a7bb1956bf80",
"key": "prop-independently-rejected-claims-no-credit",
"scope": "The accepted EM-0026 independent semantic review.",
"text": "Nine claim occurrences were independently rejected because their captured spans did not entail the complete asserted proposition.",
"visibility": "public"
},
{
"id": "em:dossier-proposition:sha256:e4fd81bf270543f9e7628c962f2b6c2907004902c82595217bf62d570a5b66b9",
"key": "prop-relation-derived-counts",
"scope": "Counts derived from the accepted EM-0026 relation ledger.",
"text": "The accepted packet contains 8 reports, 48 citation occurrences, 30 cited URLs, 27 resolving URL roots, 11 source works, 14 examined editions, 72 matched exact-span roots, 7 candidate warrant roots, 34 unresolved citations, and 20 unsupported or force-raised claim occurrences.",
"visibility": "public"
},
{
"id": "em:dossier-proposition:sha256:913134ebeedb4a61c33c6b4c7a61bebb59b5be78daf7424f26f5b14f56ba8283",
"key": "prop-keplinger-metadata-versus-support-73dfc141c5",
"scope": "Review status: candidate-supported-with-scope; data root data:keplinger-three-run-five-system-audit; method root method:human-reference-and-sentence-concordance-audit; derivation root derivation:keplinger-published-tables-and-figure.",
"text": "In one five-system dermatology-review audit, reference identifiability and metadata correctness did not establish sentence-level claim-citation concordance.",
"visibility": "public"
},
{
"id": "em:dossier-proposition:sha256:b476681a6c3ad4658a9fa593caeb6d01207a48f1b4a0fcfd4b9523dfd8dcf46d",
"key": "prop-deepresearch-bench-fact-ab2a1c0fd8",
"scope": "Review status: candidate-supported-with-edition-boundary; data root data:deepresearch-bench-100-tasks-and-system-outputs; method root method:fact-statement-url-support-evaluator; derivation root derivation:deepresearch-bench-edition-specific-tables.",
"text": "DeepResearch Bench evaluated binary support for deduplicated statement-URL pairs; numeric values are edition-specific.",
"visibility": "public"
},
{
"id": "em:dossier-proposition:sha256:06007bbafd55f21604b29a9b70bc781072fe39dfd6bd9078e539581ef49200bb",
"key": "prop-deeptrace-support-variation-de1dfdaff3",
"scope": "Review status: candidate-supported-except-disputed-gemini-number; data root data:deeptrace-303-questions-2727-outputs; method root method:deeptrace-statement-source-matrix; derivation root derivation:deeptrace-tables-and-prose.",
"text": "DeepTRACE measured statement-source support and found between-system variation; its Gemini value is internally inconsistent between table and prose.",
"visibility": "public"
},
{
"id": "em:dossier-proposition:sha256:e677a533523393e22cecf616372c9dc369ac1c4e455a851917672f69f59bb4ea",
"key": "prop-cnv-link-relevance-support-gap-81b86b6d92",
"scope": "Review status: candidate-supported-with-upstream-dependence; data root data:cnv-130-drbench-and-browsecomp-queries; method root method:cnv-ast-access-relevance-fact-check; derivation root derivation:cnv-main-table.",
"text": "Cited but Not Verified operationalized link access, topical relevance, and factual support separately and reported materially different rates.",
"visibility": "public"
},
{
"id": "em:dossier-proposition:sha256:3c8f5ea721521491b9da76d1bafc7bf2055194ff318c2f3a3dc34ccfd3f3bff6",
"key": "prop-cnv-search-depth-ablation-dad05fa3f9",
"scope": "Review status: candidate-supported-as-within-harness-association; data root data:cnv-depth-ablation-two-model-setups; method root method:cnv-ast-access-relevance-fact-check; derivation root derivation:cnv-depth-ablation-tables.",
"text": "Within the Cited but Not Verified harness, the 2-to-150-call ablation reduced reported Fact Check scores for two setups without establishing a general causal law about search depth.",
"visibility": "public"
},
{
"id": "em:dossier-proposition:sha256:06ffd95abe646fb91cec30d455f3bbb9fb44f6b410d1536060e4051a3ddf6b0b",
"key": "prop-url-health-resolution-9fcccc534a",
"scope": "Review status: candidate-supported-with-semantic-exclusion; data root data:url-health-drbench-and-expertqa-urls; method root method:http-resolution-wayback-classification; derivation root derivation:url-health-observational-results.",
"text": "The URL-health study measured HTTP resolution and Wayback presence, not semantic claim support, on outputs including reused DeepResearch Bench material.",
"visibility": "public"
},
{
"id": "em:dossier-proposition:sha256:847cf057bf61cafb9f3f9d6041e64b87cb3e7ea069664e6bf92040ad66136b25",
"key": "prop-url-health-correction-loop-f63084450b",
"scope": "Review status: pending-insufficient-semantic-span-closure; data root data:url-health-correction-experiment; method root method:url-checking-correction-loop; derivation root derivation:url-health-correction-results.",
"text": "The URL-health paper's bounded correction loop reduced non-resolving URLs but did not evaluate surviving claim support.",
"visibility": "public"
},
{
"id": "em:dossier-proposition:sha256:9ad38ce48023866a877ebd02b5af9e70e587f22da770e87e0806d9f284351aab",
"key": "prop-reportbench-match-rate-46ddf0a876",
"scope": "Review status: candidate-supported-with-judge-limitation; data root data:reportbench-100-survey-tasks; method root method:reportbench-statement-source-semantic-match; derivation root derivation:reportbench-table-1.",
"text": "ReportBench compared cited statements with retrieved cited-page content and reported sub-100-percent semantic match rates in its bounded survey-task setting.",
"visibility": "public"
},
{
"id": "em:dossier-proposition:sha256:177a37d1652c9c11537c2ff240613f1f73e9c9f1328938cb510535e3f76cfea6",
"key": "prop-researcherbench-faithfulness-groundedness-0deb7adf49",
"scope": "Review status: pending-insufficient-semantic-span-closure; data root data:researcherbench-65-frontier-ai-questions; method root method:researcherbench-faithfulness-groundedness; derivation root derivation:researcherbench-table-2.",
"text": "ResearcherBench separated support precision among cited claims from citation coverage across factual claims.",
"visibility": "public"
},
{
"id": "em:dossier-proposition:sha256:4b11d12b4521052997db748829e7d7f079024a9e2dc5adabbc2d19968469faee",
"key": "prop-liveresearchbench-e1-e2-e3-07e84c4837",
"scope": "Review status: pending-insufficient-semantic-span-closure; data root data:liveresearchbench-wide-info-and-market-analysis; method root method:liveresearchbench-e1-e2-e3-rubric-tree; derivation root derivation:liveresearchbench-v5-table-7.",
"text": "LiveResearchBench v5 separately typed inaccessible/non-resolving, irrelevant, and unsupported-claim citation errors for two task categories; its table reports per-report errors without raw denominators.",
"visibility": "public"
},
{
"id": "em:dossier-proposition:sha256:df256ad7c8cee697080f219ef3afa64d990e76522da9566eea197ee35a20d7c8",
"key": "prop-citation-verifier-calibration-da47a19a65",
"scope": "Review status: pending-insufficient-semantic-span-closure; data root data:citationverbench-1248-human-reviewed-decisions; method root method:citationverbench-rubric-judge-comparison; derivation root derivation:citationverbench-reported-calibration.",
"text": "A human-reviewed verifier benchmark showed that rubric-LLM citation-support judgments retain model-dependent error and directional bias.",
"visibility": "public"
}
],
"question": "What empirical evidence published or publicly posted by 2026-08-22 measures whether citations produced by deep-research agents resolve and actually support the claims made from them?",
"scope": "Eight context-isolated public-by-design reports captured under one frozen prompt, plus the public source editions and exact spans they cited. This historical pilot does not estimate current or universal agent behavior.",
"source_works": [
{
"canonical_uri": "https://arxiv.org/html/2607.08700v1",
"creators": [
"Ethan Leung",
"Elias Lumer",
"Corey Feld",
"Austin Huber",
"Vamse Kumar Subbiah",
"Kevin Paul"
],
"id": "em:dossier-source-work:sha256:22b2290799937200bec1536c9c67de0f5642340c180e656e8ed687a1d566f8d9",
"key": "work-citation-verifier-benchmark-2d5e94336b",
"kind": "paper",
"license": "CC BY 4.0",
"title": "Do You Need a Frontier Model as a Citation Verifier? Benchmarking Rubric LLMs for Deep-Research Source Attribution",
"visibility": "public"
},
{
"canonical_uri": "https://arxiv.org/html/2605.06635v1",
"creators": [
"Hailey Onweller",
"Elias Lumer",
"Austin Huber",
"Pia Ramchandani",
"Vamse Kumar Subbiah",
"Corey Feld"
],
"id": "em:dossier-source-work:sha256:19d52572cf9e426b68cba35f9dd627d014ff71a978756c5b50103c2aabe554ce",
"key": "work-cited-not-verified-3825b25622",
"kind": "paper",
"license": "CC BY 4.0",
"title": "Cited but Not Verified: Parsing and Evaluating Source Attribution in LLM Deep Research Agents",
"visibility": "public"
},
{
"canonical_uri": "https://arxiv.org/html/2506.11763v1",
"creators": [
"Mingxuan Du",
"Benfeng Xu",
"Chiwei Zhu",
"Licheng Zhang",
"Xiaorui Wang",
"Zhendong Mao"
],
"id": "em:dossier-source-work:sha256:2a2c79463c9697b114728426edb183b5881f8b742ea6febfc272f73db58c1a14",
"key": "work-deepresearch-bench-paper-8ef36010d2",
"kind": "paper",
"license": "CC BY 4.0; unknown",
"title": "DeepResearch Bench: A Comprehensive Benchmark for Deep Research Agents",
"visibility": "public"
},
{
"canonical_uri": "https://github.com/Ayanami0730/deep_research_bench/commit/469cce54ea7f6a63c163d3d9fec879cf289ec484",
"creators": [
"DeepResearch Bench project"
],
"id": "em:dossier-source-work:sha256:433604e28e4a1ee90dfc7f2c5b2389d04e64a600930ad51cd9acf599b4df9f21",
"key": "work-deepresearch-bench-repository-1e475c7631",
"kind": "webpage",
"license": "Apache-2.0",
"title": "Ayanami0730/deep_research_bench",
"visibility": "public"
},
{
"canonical_uri": "https://arxiv.org/html/2509.04499v1",
"creators": [
"Pranav Narayanan Venkit",
"Philippe Laban",
"Yilun Zhou",
"Kung-Hsiang Huang",
"Yixin Mao",
"Chien-Sheng Wu"
],
"id": "em:dossier-source-work:sha256:f58addec09e622c4c6457ae395e065ad2c0e9bc92035131100c8cb15e3a9d6d1",
"key": "work-deeptrace-650d66cfcb",
"kind": "paper",
"license": "arXiv non-exclusive distribution license; unknown",
"title": "DeepTRACE: Auditing Deep Research AI Systems for Tracking Reliability Across Citations and Evidence",
"visibility": "public"
},
{
"canonical_uri": "https://pmc.ncbi.nlm.nih.gov/articles/PMC13109748/",
"creators": [
"Lauren E. Keplinger",
"Luke K. Frashure",
"Sabrina A. Duran",
"Gangqing Hu"
],
"id": "em:dossier-source-work:sha256:7e292f6ce412d254983ada9a5032ef5fac626876b16404d99ed1a99ce4e3ba05",
"key": "work-keplinger-dermatology-audit-639f9ea37a",
"kind": "paper",
"license": "CC BY-NC 4.0",
"title": "Assessment of Deep Research for dermatology literature reviews: Deep concern over the hype",
"visibility": "public"
},
{
"canonical_uri": "https://data.mendeley.com/datasets/3s73z9zf3c/1",
"creators": [
"Gangqing Hu",
"West Virginia University"
],
"id": "em:dossier-source-work:sha256:daa2447b3107ce9818b1c435de19c6ea8ca0de07129761714bda29dc7a062eb4",
"key": "work-keplinger-supplement-1a03034fb3",
"kind": "dataset",
"license": "CC BY 4.0",
"title": "Supplementary materials of the article: Assessment of Deep Research for Dermatology Literature Reviews: Deep Concern Over the Hype",
"visibility": "public"
},
{
"canonical_uri": "https://arxiv.org/pdf/2510.14240v5",
"creators": [
"Jiayu Wang",
"Yifei Ming",
"Riya Dulepet",
"Qinglin Chen",
"Austin Xu",
"Zixuan Ke",
"Frederic Sala",
"Aws Albarghouthi",
"Caiming Xiong",
"Shafiq Joty"
],
"id": "em:dossier-source-work:sha256:f93da6e764e8998aec2b860736cdf95f0218d6144a5827faf13b95db0f5720fd",
"key": "work-liveresearchbench-61e7da557b",
"kind": "paper",
"license": "CC BY-NC-SA 4.0",
"title": "LiveResearchBench: A Live Benchmark for User-Centric Deep Research in the Wild",
"visibility": "public"
},
{
"canonical_uri": "https://arxiv.org/html/2508.15804v1",
"creators": [
"Minghao Li",
"Ying Zeng",
"Zhihao Cheng",
"Cong Ma",
"Kai Jia"
],
"id": "em:dossier-source-work:sha256:a3c37defe96fa413eeb6391ba84067bec7456d35921aaf1b38c7ad003078fb54",
"key": "work-reportbench-dca823c910",
"kind": "paper",
"license": "CC BY 4.0",
"title": "ReportBench: Evaluating Deep Research Agents via Academic Survey Tasks",
"visibility": "public"
},
{
"canonical_uri": "https://arxiv.org/html/2507.16280v1",
"creators": [
"ResearcherBench authors",
"SII-GAIR"
],
"id": "em:dossier-source-work:sha256:8ece2d6fcbf199ebb71533abd1448ccd8796f710bb7759b59557d9856a17a2a3",
"key": "work-researcherbench-82729ab60f",
"kind": "paper",
"license": "arXiv non-exclusive distribution license",
"title": "ResearcherBench: Evaluating Deep AI Research Systems on the Frontiers of Scientific Inquiry",
"visibility": "public"
},
{
"canonical_uri": "https://arxiv.org/html/2604.03173v1",
"creators": [
"Delip Rao",
"Eric Wong",
"Chris Callison-Burch"
],
"id": "em:dossier-source-work:sha256:ebddaffccfcd23869d5dd243d9f4272b520252f96cd784ed4058bacef7f764aa",
"key": "work-url-health-f64342d493",
"kind": "paper",
"license": "CC0 1.0",
"title": "Detecting and Correcting Reference Hallucinations in Commercial LLMs and Deep Research Agents",
"visibility": "public"
},
{
"canonical_uri": "https://github.com/yoheinakajima/epistemedia/tree/main/research/how-we-know/agent-citation-lineage",
"creators": [
"Epistemedia EM-0026 research process"
],
"id": "em:dossier-source-work:sha256:c552ac33696c6615477f161b374123595dd33681366befc990cbee17731e6e8b",
"key": "work-em0026-audit-instrument",
"kind": "instrument",
"license": "Repository metadata and derived audit relations under Apache-2.0; embedded source excerpts retain their recorded treatments",
"title": "EM-0026 deterministic agent-citation evidence ledger",
"visibility": "public"
}
],
"spans": [
{
"digest": "sha256:8bebee03d81b15ad0af5b8ef69993f7e3cf525a7daa6e2608007fa9bac9b9a40",
"edition_key": "edition-citation-verifier-arxiv-v1-acfd4abab7",
"extent": {
"type": "json-value",
"value": "624 attribution-citation pairs with gold labels for all 1,248 LLM-judged decisions, every one human-reviewed"
},
"id": "em:dossier-span:sha256:c2f13c8a50d0c4d694de87dd111bc046e5b64061f3481f1cf4cb41878791b099",
"key": "span-6885cf2524462022475b5829da081b2d35fcfe7463ca6ad0-3b1a122faf",
"locator": {
"label": "Section 1 contributions",
"pointer": "/excerpts/0/text",
"type": "json-pointer"
},
"visibility": "public"
},
{
"digest": "sha256:6a7d5dd08b03dbb3fb7acaa3d8f8943e1ee7d8dcaeb8b2b829ba36a7963e4991",
"edition_key": "edition-citation-verifier-arxiv-v1-acfd4abab7",
"extent": {
"type": "json-value",
"value": "378 of which were hard cases"
},
"id": "em:dossier-span:sha256:1628933030cf1849267cc250802d96ad06a1dc463ebc598b3034885f5eefcd3f",
"key": "span-794bdf7e0595799afbb0d55f3dc522077aed9615c59f1e35-43f2ba944d",
"locator": {
"label": "Abstract, line 16",
"pointer": "/excerpts/1/text",
"type": "json-pointer"
},
"visibility": "public"
},
{
"digest": "sha256:0e980e624cddfead247a93771e68f3c678bd30495cd310fd81c0812ecab2bb8b",
"edition_key": "edition-citation-verifier-arxiv-v1-acfd4abab7",
"extent": {
"type": "json-value",
"value": "1,248 rubric decisions"
},
"id": "em:dossier-span:sha256:df4fc5c1b77f89b40d53711554081c09d64b3257a54bf32e74be00b3f15d9ca4",
"key": "span-e54ee1457854a9a5e46f7c7c2e66ac6a5a69b963f8ee3450-8f875cfe94",
"locator": {
"label": "Abstract, line 16",
"pointer": "/excerpts/2/text",
"type": "json-pointer"
},
"visibility": "public"
},
{
"digest": "sha256:e29185e1dd93d69d1cf4328288773206e15b35990c7de068b31765e058bd50fd",
"edition_key": "edition-cnv-arxiv-v1-f24001c50a",
"extent": {
"type": "json-value",
"value": "from 79% to 17% (62%)"
},
"id": "em:dossier-span:sha256:f70e443b9842d25a613f4ee0665344914a109815031eda205da5cd0babba10ae",
"key": "span-417182b4c15ac550469e435a02aab1e91b395f72f186312d-e9e2c4a70c",
"locator": {
"label": "Section 4.3, line 212",
"pointer": "/excerpts/0/text",
"type": "json-pointer"
},
"visibility": "public"
},
{
"digest": "sha256:af0b066b292e6d887067c6cfa9802eb8b1984929b622d91cdca61d0e8a3bc31f",
"edition_key": "edition-cnv-arxiv-v1-f24001c50a",
"extent": {
"type": "json-value",
"value": "the LLM-as-a-judge approach ... may retain biases inherent to the judge model"
},
"id": "em:dossier-span:sha256:c488019e611ca8dbe62a30e922a1fc7f2201fc1f5caedd8da2886de924d82e35",
"key": "span-5cd9ba4ca72da10cf51c0beddc6ea2765c74721bdd526a67-ed8f789629",
"locator": {
"label": "Limitations, lines 219-224",
"pointer": "/excerpts/1/text",
"type": "json-pointer"
},
"visibility": "public"
},
{
"digest": "sha256:4046b3ea7b246aa51b11f554509b50ebde3ca883cdcb8424e74a663b354bb845",
"edition_key": "edition-cnv-arxiv-v1-f24001c50a",
"extent": {
"type": "json-value",
"value": "Fact Check accuracy drops approximately 42% on average from minimal (2 calls) to maximal search depth"
},
"id": "em:dossier-span:sha256:48368b4a9e46339c88189de3ff152e4048ed062992fb4789c77e47e645bcc865",
"key": "span-6e961498ed0f81a950f775956587f741e0eb9f11640fd3cb-afa5b59854",
"locator": {
"label": "Section 4.3, Tables 2–3",
"pointer": "/excerpts/2/text",
"type": "json-pointer"
},
"visibility": "public"
},
{
"digest": "sha256:a4386d56325583893d806fdd15ccd475d5096f0b445b1b9f52aeba253c910987",
"edition_key": "edition-cnv-arxiv-v1-f24001c50a",
"extent": {
"type": "json-value",
"value": "\"Fact Check verifies whether specific factual claims are accurately supported by the source content.\""
},
"id": "em:dossier-span:sha256:d4e2f35392b064365ad867f33ead8636bb8bbf3af03c0141585e5ae56c868090",
"key": "span-732c03e69c6e142a492275cee5d1e2d474c75422c9b84aaa-f7675998cc",
"locator": {
"label": "Section 3.3.1-3.3.3",
"pointer": "/excerpts/3/text",
"type": "json-pointer"
},
"visibility": "public"
},
{
"digest": "sha256:be05510d58cc0231bf968f29c70b3d669a82d5cd0c74a778a5f8c4e035bf45b4",
"edition_key": "edition-cnv-arxiv-v1-f24001c50a",
"extent": {
"type": "json-value",
"value": "yet achieve only 39–77% factual accuracy"
},
"id": "em:dossier-span:sha256:ea33bc17365ed01ed68ab06f33c4fd2020dc8e6947659e4e9629b44af4aef635",
"key": "span-74f6acafbb4c54d556a572460bf1b45bacd1e293f4aa16cc-5ee15c82c4",
"locator": {
"label": "Abstract, lines 47-48",
"pointer": "/excerpts/4/text",
"type": "json-pointer"
},
"visibility": "public"
},
{
"digest": "sha256:44d4c6941f2920e907de1e4cd22aeb9a725f24cc19432723754e316bd6f5cc49",
"edition_key": "edition-cnv-arxiv-v1-f24001c50a",
"extent": {
"type": "json-value",
"value": "link validity above 94% and relevance above 80%, yet achieve only 39–77% factual accuracy"
},
"id": "em:dossier-span:sha256:49ee66722d6d523dd485f05c796181af55736c3ceab2a235c5c3b40cca95e439",
"key": "span-912c931063109f84ea88aea34891dd5e0f9147d2176af258-4d566ca076",
"locator": {
"label": "Abstract and Table 1",
"pointer": "/excerpts/5/text",
"type": "json-pointer"
},
"visibility": "public"
},
{
"digest": "sha256:e35ca66d12349d7290a6dbe7449bf6da173eb0671765644b581713528b316c74",
"edition_key": "edition-cnv-arxiv-v1-f24001c50a",
"extent": {
"type": "json-value",
"value": "maintain link validity above 94% and relevance above 80%, yet achieve only 39–77% factual accuracy"
},
"id": "em:dossier-span:sha256:8cf2ff4c32b2dac6d6dfc7684dd2bf7dca9d40fe7b7b36d50dfdcddad067bff3",
"key": "span-bb16d5f06abe8641d9ea6694e549aa15e240dd88946aa47a-5946b6e6b7",
"locator": {
"label": "Abstract and Table 1",
"pointer": "/excerpts/6/text",
"type": "json-pointer"
},
"visibility": "public"
},
{
"digest": "sha256:e084f6ff65e9a116b127ab06d9745be6354fbcf3e66052c0182bc38177a9ae48",
"edition_key": "edition-cnv-arxiv-v1-f24001c50a",
"extent": {
"type": "json-value",
"value": "contradicted, absent, or uncertain"
},
"id": "em:dossier-span:sha256:5b052dcb433428d39db474fc5c8ae7549944ea380839cf60538b66940163df78",
"key": "span-c3dfc98566c27561eb7267972e00e3e1ba6a399a3ff5d002-37f40d3260",
"locator": {
"label": "page 5, Section 3.3.3",
"pointer": "/excerpts/7/text",
"type": "json-pointer"
},
"visibility": "public"
},
{
"digest": "sha256:a14ffeb203983bb9ab414d79300766c78a29ade822b77d94bdd51c6788cfe450",
"edition_key": "edition-cnv-arxiv-v1-f24001c50a",
"extent": {
"type": "json-value",
"value": "\"LLM-as-a-judge ... may retain biases inherent to the judge model\""
},
"id": "em:dossier-span:sha256:cd2b068589af7da7601362a457b7ac96058de3989dd5187d9e11e3e532e29699",
"key": "span-c7dbd1d9e09980703ecbee8161594e4b426c1af63117b286-153d26f046",
"locator": {
"label": "Section 5 'Limitations'",
"pointer": "/excerpts/8/text",
"type": "json-pointer"
},
"visibility": "public"
},
{
"digest": "sha256:d34096a9d9b1fbb43ef50ba503c3894607c008b563e1e1a5d1643cbc23c5d447",
"edition_key": "edition-cnv-arxiv-v1-f24001c50a",
"extent": {
"type": "json-value",
"value": "Citations are evaluated along three dimensions. (1) Link Works verifies URL accessibility, (2) Relevant Content measures topical alignment, and (3) Fact Check validates factual accuracy against source content."
},
"id": "em:dossier-span:sha256:237508b682de825711f61762deb5ce7661d5d197e9fde3aed5f5e80f05fdcdb5",
"key": "span-cd21c450c0d33df20a2b1a539d5725347db1b5bab7d1e986-2e3e0c3ac7",
"locator": {
"label": "Abstract, lines 45-49",
"pointer": "/excerpts/9/text",
"type": "json-pointer"
},
"visibility": "public"
},
{
"digest": "sha256:4ff735800b0f2aca92c4c6865a7bcf5746c1b7711a6105d5403bdffb68adce63",
"edition_key": "edition-cnv-arxiv-v1-f24001c50a",
"extent": {
"type": "json-value",
"value": "\"Fact Check accuracy drops approximately 42% on average\""
},
"id": "em:dossier-span:sha256:354ceda8de2c156cc583b4cbb66b6a0f60199ddba0461331a5c1b121e09c2000",
"key": "span-d638c8b21f0e4a0f840880b6ffe0217ce1623758cc09fdcc-8585f6d39b",
"locator": {
"label": "Section 4.3, Tables 2-3 and following paragraph",
"pointer": "/excerpts/10/text",
"type": "json-pointer"
},
"visibility": "public"
},
{
"digest": "sha256:da34a8e2a1d4e190d3b2095d79886b3cbf7edb2c63874189aa381302d2c2924e",
"edition_key": "edition-cnv-arxiv-v1-f24001c50a",
"extent": {
"type": "json-value",
"value": "only 1 failed link out of 2,159 evaluations"
},
"id": "em:dossier-span:sha256:2bd67a5caa606894f4301ca1185c88a4dde02d3c5c4a2cf311463a0f39332443",
"key": "span-deada8bacb987a47f2a42e52fbfeb5c1beb3d250c0be79e2-0e91f5cacc",
"locator": {
"label": "Section 4.4, line 217",
"pointer": "/excerpts/11/text",
"type": "json-pointer"
},
"visibility": "public"
},
{
"digest": "sha256:a0eeca5c86f9c7aeda9167f626a8cb0ed56dc87ea1ac5ca8c4a7a0add76e7625",
"edition_key": "edition-deeptrace-arxiv-v1-70a6e720c1",
"extent": {
"type": "json-value",
"value": "citation accuracy ranging from 40–80% across systems"
},
"id": "em:dossier-span:sha256:e748230bbc2f6b31433eaba4536917834c454cb73a7f18e8d9b975e2cac447b5",
"key": "span-8cce3d9ee1e1674ba1b93321dd9b99a55d834ca980451f8a-d2fceafe11",
"locator": {
"label": "Abstract, line 46",
"pointer": "/excerpts/0/text",
"type": "json-pointer"
},
"visibility": "public"
},
{
"digest": "sha256:5cc83ab699cb1342475b407e34d33fc106d64b5d9270e5bb9445d80586ebabc2",
"edition_key": "edition-deeptrace-arxiv-v1-70a6e720c1",
"extent": {
"type": "json-value",
"value": "303 queries x 9 models"
},
"id": "em:dossier-span:sha256:8c1ac40e63a414c1a41cd0842632677269c8800967f22ef8247b6fb36f6242ea",
"key": "span-9dd6eb46529e77a62748093528e52685fbd92d46a756aef7-96d75d9e2b",
"locator": {
"label": "Section 3.2, lines 163-171",
"pointer": "/excerpts/1/text",
"type": "json-pointer"
},
"visibility": "public"
},
{
"digest": "sha256:bb4f28cd0f4527c633bcfacf51b18ced7387c4a2f266c59d41de7ee899b24b85",
"edition_key": "edition-deeptrace-iclr-2026-b2b153eaa6",
"extent": {
"type": "json-value",
"value": "Factual support (statement-source) 0.62 binary"
},
"id": "em:dossier-span:sha256:6cf2883b02dd404844a662bae3367ab60d91f6847160a3df9c6ec6b287dc0858",
"key": "span-35927851edd2e7e5e0f60d98498940f88304ba99bf1f85a0-2fe586436b",
"locator": {
"label": "page 14, Table 3",
"pointer": "/excerpts/0/text",
"type": "json-pointer"
},
"visibility": "public"
},
{
"digest": "sha256:2ce91759797b86455a478a9ced58ec59d032e1a32a1c964347f96be4ab9e2054",
"edition_key": "edition-deeptrace-iclr-2026-b2b153eaa6",
"extent": {
"type": "json-value",
"value": "\"The dataset comprises 303 questions\""
},
"id": "em:dossier-span:sha256:b9cbe6b1d35e74f3c8133d2be41a47d9361a26958e9f644dafd97e9e5acb516a",
"key": "span-9cb1700c08368feb234207045354cbf559d6ed80f2b0bae8-f898722eb5",
"locator": {
"label": "pp.4-7, Sections 3.1.1 and 3.2",
"pointer": "/excerpts/1/text",
"type": "json-pointer"
},
"visibility": "public"
},
{
"digest": "sha256:af76864a50c641f8d693216d3b7ebe2a414eebf75467305c1e1a6946b6746590",
"edition_key": "edition-deeptrace-iclr-2026-b2b153eaa6",
"extent": {
"type": "json-value",
"value": "with citation accuracy ranging from 40–80% across systems"
},
"id": "em:dossier-span:sha256:06ea026aae360f700369f4b1f658c4ca85e8acce431923cd7978ff2358b1ec05",
"key": "span-a08468cb62f2feb21d9cbb95175ce00f54592bd83573cc91-03f65cfa45",
"locator": {
"label": "Official ICLR abstract",
"pointer": "/excerpts/2/text",
"type": "json-pointer"
},
"visibility": "public"
},
{
"digest": "sha256:a38c9ac275570f2b78dd1e1bf7663579e35f9037131e12d721dfb4c6f005683d",
"edition_key": "edition-deeptrace-iclr-2026-b2b153eaa6",
"extent": {
"type": "json-value",
"value": "For roughly 15% of the URLs, the Reader tool returns an error"
},
"id": "em:dossier-span:sha256:12ab62355d178256bbaed6cdf5cbf6ae150410d9364f98cced347b54161fa5e8",
"key": "span-a15ab02e9fcb23df02fb03d28967d4220ddf65ad380d2230-ddf85fa359",
"locator": {
"label": "page 4, source scraping",
"pointer": "/excerpts/3/text",
"type": "json-pointer"
},
"visibility": "public"
},
{
"digest": "sha256:75a9212dc448c15ffc1e4a36bdf190728977470661bd9f45be534010704faa25",
"edition_key": "edition-deeptrace-iclr-2026-b2b153eaa6",
"extent": {
"type": "json-value",
"value": "\"Factual support (statement–source) 0.62\""
},
"id": "em:dossier-span:sha256:2aa386b3653a99db3b670f783272056371480ed287dc1f62a51d3d587d680996",
"key": "span-aa1534ce2e177153ab85e512deb8d064d14ec1eae16d8032-2299096a09",
"locator": {
"label": "p.14, Table 3",
"pointer": "/excerpts/4/text",
"type": "json-pointer"
},
"visibility": "public"
},
{
"digest": "sha256:d2f6a4561a938ee913dadde2fc4a80c32c492fbcb3b47f236c9639375843ae8d",
"edition_key": "edition-deeptrace-iclr-2026-b2b153eaa6",
"extent": {
"type": "json-value",
"value": "Pearson correlation of 0.62 between the LLM judge and manual labels"
},
"id": "em:dossier-span:sha256:a575506a3cca26acda376aa922fdb0e269767f71b01f502a7c50f3d4492e6f03",
"key": "span-e2c9f6220843a04665f5d1cd142ac15f04211ace5c579fa6-c5e0476df0",
"locator": {
"label": "Paper section 3.1.1, factual-support validation",
"pointer": "/excerpts/5/text",
"type": "json-pointer"
},
"visibility": "public"
},
{
"digest": "sha256:cf32a686cec566152749e3a4c643fa29aa115099f9eba77c0ca113c48264f769",
"edition_key": "edition-deeptrace-iclr-2026-b2b153eaa6",
"extent": {
"type": "json-value",
"value": "Gemini(DR) demonstrates weak citation performance: only 40.3% citation accuracy"
},
"id": "em:dossier-span:sha256:533550e262b8691ce9ed33aa0e89a7bf13d89c04c6002e1855ed1e767f556e3a",
"key": "span-supplement-deeptrace-gemini-prose-40-3-ed29752a6f",
"locator": {
"label": "printed page 10; PDF page index 10/22; Section 4 Results; Deep Research Agents paragraph",
"pointer": "/review_supplements/0/extent",
"type": "json-pointer"
},
"visibility": "public"
},
{
"digest": "sha256:e020a09063eee7aa1eb70663cf3e8bd7a8162952f5ae65d959d57ef74ab855f6",
"edition_key": "edition-deeptrace-iclr-2026-b2b153eaa6",
"extent": {
"type": "json-value",
"value": {
"cell": "50.3",
"column": "Gemini (DR)",
"row": "%Citation Accuracy"
}
},
"id": "em:dossier-span:sha256:cf4e063e6f297fe901b342752b67850d84cc3e8e21ace2ef8d5ea4f3fd93091c",
"key": "span-supplement-deeptrace-gemini-table-50-3-7eb7bd254b",
"locator": {
"label": "printed page 9; PDF page index 9/22; Table 1; Gemini (DR) column; %Citation Accuracy row",
"pointer": "/review_supplements/1/extent",
"type": "json-pointer"
},
"visibility": "public"
},
{
"digest": "sha256:1bde849b0f5df91e847e907dd763f27be01ffdc7eb0c7e61898c8a9731472c35",
"edition_key": "edition-drbench-arxiv-v1-7afa5ade36",
"extent": {
"type": "json-value",
"value": "Each unique Statement-URL pair undergoes a support evaluation."
},
"id": "em:dossier-span:sha256:ed08815f4aabd455caf37f61a1e37a2dc35725098f2e9be5c6da6c9d5854d612",
"key": "span-015bb68c904bf225e84c82b1acffb3cc38a4a087d7b0f79d-4c454092b1",
"locator": {
"label": "Evaluation methodology, lines 144-149",
"pointer": "/excerpts/0/text",
"type": "json-pointer"
},
"visibility": "public"
},
{
"digest": "sha256:53d38f9cf0d0aedf28a77d7de32935372202fff78a0c02d98f8c269ef223906a",
"edition_key": "edition-drbench-arxiv-v1-7afa5ade36",
"extent": {
"type": "json-value",
"value": "Perplexity Deep Research | ... | 90.24 | 31.26 ... Gemini-2.5-Pro Deep Research | ... | 81.44 | 111.21 ... OpenAI Deep Research | ... | 77.96 | 40.79"
},
"id": "em:dossier-span:sha256:fd8835d2aaa9358bd7dfbcf07848ed19001fa82f11e45ceff3374cdd4e882be3",
"key": "span-0d4f357f2a5565bdaea01a1e537650958d71216868d1a2b0-9b0af35d76",
"locator": {
"label": "Table 1, lines 177-182",
"pointer": "/excerpts/1/text",
"type": "json-pointer"
},
"visibility": "public"
},
{
"digest": "sha256:a49c200ea10833496ac73719992be09d5fd3fd021fdd51ccc587ac2592225f56",
"edition_key": "edition-drbench-arxiv-v1-7afa5ade36",
"extent": {
"type": "json-value",
"value": "‘support’ or ‘not support’"
},
"id": "em:dossier-span:sha256:4648142e8019a2d54f201844685fbb87a921d93fc24b0f5c08f8124a82ffc374",
"key": "span-13cb0ddb087b82988b7445a9cfb02562f9c044ef4f244e8b-87c3b42c68",
"locator": {
"label": "Section 3.2, lines 144-146",
"pointer": "/excerpts/2/text",
"type": "json-pointer"
},
"visibility": "public"
},
{
"digest": "sha256:64c911242980826f92ffae6d0571646ecc3aceca73c213791c6d3a8d889b4cc7",
"edition_key": "edition-drbench-arxiv-v1-7afa5ade36",
"extent": {
"type": "json-value",
"value": "90.24"
},
"id": "em:dossier-span:sha256:b118f44725ec505886f92a94123fe289be763523b8afb669757c72ccd0d8c352",
"key": "span-196c54d6212fbc0c054dbf8e8467f2388c1f788bb93f4b1a-d460e927f6",
"locator": {
"label": "Table 1, Perplexity Deep Research citation-accuracy cell",
"pointer": "/excerpts/3/text",
"type": "json-pointer"
},
"visibility": "public"
},
{
"digest": "sha256:dedd156999fa7d57ae9c22a166c56c263547eee3c37e6f41fe7644d980110986",
"edition_key": "edition-drbench-arxiv-v1-7afa5ade36",
"extent": {
"type": "json-value",
"value": "OpenAI Deep Research | 46.98 | 46.87 | 45.25 | 49.27 | 47.14 | 77.96 | 40.79"
},
"id": "em:dossier-span:sha256:762ebe0afa7a2ddf8eb19d879c91f63d1de7d25192f10d248b65fefb0e113038",
"key": "span-2f311bfb451c6208101cb7583af37c338914fcdb10f5c75b-90e7764ce2",
"locator": {
"label": "Table 1, OpenAI Deep Research row; FACT columns",
"pointer": "/excerpts/4/text",
"type": "json-pointer"
},
"visibility": "public"
},
{
"digest": "sha256:86d5d20702fbe736e87099be370fadf16b4acea42bea16ba28ebb473ba2ae670",
"edition_key": "edition-drbench-arxiv-v1-7afa5ade36",
"extent": {
"type": "json-value",
"value": "\"OpenAI Deep Research | April 1 – May 8\""
},
"id": "em:dossier-span:sha256:e247bc28c78a353c87489473329585b1051cf199a4b66c81569e31947e10386a",
"key": "span-76e9bd59b31c7ebd2ed7a2186e59753a79d3c730e0d7b4fb-9c98d99242",
"locator": {
"label": "Appendix D, Table 5",
"pointer": "/excerpts/5/text",
"type": "json-pointer"
},
"visibility": "public"
},
{
"digest": "sha256:3cca518ed14d7451346a35623b20e7f9e350e4cb5cbcec3d7f622cc6aac509ce",
"edition_key": "edition-drbench-arxiv-v1-7afa5ade36",
"extent": {
"type": "json-value",
"value": "\"Each unique Statement-URL pair undergoes a support evaluation.\""
},
"id": "em:dossier-span:sha256:aba61c5b1685e20da29c0d6a42dd49b66f0b6c2f3ca08483716ae5c518562d4f",
"key": "span-9fe42cb48702d8d96ebbbc0309692214ece6ec495c447c06-80da541506",
"locator": {
"label": "Section 3.2, paragraphs 'Statement-URL Pair Extraction and Deduplication' and 'Support Judgment'",
"pointer": "/excerpts/6/text",
"type": "json-pointer"
},
"visibility": "public"
},
{
"digest": "sha256:4edb75ec3f690eb2d15089923ba5e9c4e6174f55094d0919041bf10625901ceb",
"edition_key": "edition-drbench-arxiv-v1-7afa5ade36",
"extent": {
"type": "json-value",
"value": "\"Each unique Statement-URL pair undergoes a support evaluation... This results in a binary judgment ('support' or 'not support') for each pair, determining whether the citation accurately grounds the claim.\""
},
"id": "em:dossier-span:sha256:0b565d8281beb5c9b95d300108446049bc4c93de4dc3e80a501f41e85d0a98a5",
"key": "span-b0a39c963cdfdc5f9852cfc9f7e33e05d31232bd1d899885-adf9545222",
"locator": {
"label": "p. 4, lines 150-159",
"pointer": "/excerpts/7/text",
"type": "json-pointer"
},
"visibility": "public"
},
{
"digest": "sha256:ae845f85955fa47d52d51f32803e6a26afa8b725c523e53d9ef30b2abca72d9b",
"edition_key": "edition-drbench-arxiv-v1-7afa5ade36",
"extent": {
"type": "json-value",
"value": "96% of cases"
},
"id": "em:dossier-span:sha256:d753bc9c860da325190141f3f78a0d031f2d6c7cf5e1a51f19f355f44cfb47db",
"key": "span-b83996e3b5fdf878e04d6d41d0e7a1eebee9fdd9bbccf2ae-820a3eab86",
"locator": {
"label": "Appendix C, line 365",
"pointer": "/excerpts/8/text",
"type": "json-pointer"
},
"visibility": "public"
},
{
"digest": "sha256:82b9361b6e696a92b41fdfae500f56b05d48575c0d429f15b3a8d6786c83b8d7",
"edition_key": "edition-drbench-arxiv-v1-7afa5ade36",
"extent": {
"type": "json-value",
"value": "\"the proportion of 'support' statement-URL pairs for each individual task\""
},
"id": "em:dossier-span:sha256:c6710bb4b411353c7c600e449cc556e39a3530dc63c002a7b3973ef6bb1793bf",
"key": "span-e4a9f30effc031388ed5c9993dc6f071903bec5e3f1998db-5d95269b76",
"locator": {
"label": "pp.17-18, Appendix E, equations 4-6",
"pointer": "/excerpts/9/text",
"type": "json-pointer"
},
"visibility": "public"
},
{
"digest": "sha256:3cca518ed14d7451346a35623b20e7f9e350e4cb5cbcec3d7f622cc6aac509ce",
"edition_key": "edition-drbench-arxiv-v1-7afa5ade36",
"extent": {
"type": "json-value",
"value": "\"Each unique Statement-URL pair undergoes a support evaluation.\""
},
"id": "em:dossier-span:sha256:27d667cb8cff21b646ec759c6c6fe79c5744a28079337d7a7fa9ac98abf8da9d",
"key": "span-eaf37bef7b8dc35f3e00b85d3e58ad872dc0ea604a062c8c-fa09300f85",
"locator": {
"label": "p.4, Section 3.2, lines 148-155",
"pointer": "/excerpts/10/text",
"type": "json-pointer"
},
"visibility": "public"
},
{
"digest": "sha256:e9c779296734583861b22f45da62bff61c9c899018602fda17716d69fc449a63",
"edition_key": "edition-drbench-arxiv-v1-7afa5ade36",
"extent": {
"type": "json-value",
"value": "92% of cases"
},
"id": "em:dossier-span:sha256:a18ccb20769a7f5026931d907492839678079992f3ecefcecbfe05557d211246",
"key": "span-f35c9565f2fd047f6e7272dc608bbf6a63813a717bad5354-e6a07f05b6",
"locator": {
"label": "Appendix C, line 365",
"pointer": "/excerpts/11/text",
"type": "json-pointer"
},
"visibility": "public"
},
{
"digest": "sha256:43c892b582d42a335c982009f97347b5bf2ae2799cc3e478c53504f4d01e4388",
"edition_key": "edition-drbench-arxiv-v1-7afa5ade36",
"extent": {
"type": "json-value",
"value": "\"aligned with human 'support' determinations in 96% of cases\""
},
"id": "em:dossier-span:sha256:17193f2aab7106eece041bcabda963d9bd172d798792efc058b6a93c9bd4cb0c",
"key": "span-f98f1274859ba98cc538e3be6d11d345a5382cfcb9516aa1-d60b765c48",
"locator": {
"label": "p.17, Appendix C, lines 704-708",
"pointer": "/excerpts/12/text",
"type": "json-pointer"
},
"visibility": "public"
},
{
"digest": "sha256:8be6257e74c3696813fb0079c27e93e8bec16c71d214b6da43811fbd9b406ba4",
"edition_key": "edition-drbench-iclr-2026-d00a3dcdb6",
"extent": {
"type": "json-value",
"value": "aligned with human 'support' determinations in 96% of cases"
},
"id": "em:dossier-span:sha256:a59ed2a9ea0e6972bc3f71d63feb1c361b76eb8d8fa5bb679c7beab61a32b713",
"key": "span-279db725380c133044944dccb379b75a396242219aa30437-569fc9e2fd",
"locator": {
"label": "page 16, FACT judge validation",
"pointer": "/excerpts/0/text",
"type": "json-pointer"
},
"visibility": "public"
},
{
"digest": "sha256:78f021c07980218ead835a986c84780444479e627dcfa9f2ea126b3604b89ad7",
"edition_key": "edition-drbench-iclr-2026-d00a3dcdb6",
"extent": {
"type": "json-value",
"value": "This yields a binary judgment: 'support' or 'not support'."
},
"id": "em:dossier-span:sha256:d915faef70b54c6751577ba10f4057744b2b1b0b4d10613cdd8d74720a5da23b",
"key": "span-7d7331942f9d0520e18e96206e7718e5187788be4d48cef9-dfdbc1d019",
"locator": {
"label": "page 4, Section 3.2, Support Judgment",
"pointer": "/excerpts/1/text",
"type": "json-pointer"
},
"visibility": "public"
},
{
"digest": "sha256:11034cf2238003f213bf87210c9017f57cc743910fa70754f3e61f875abe75e6",
"edition_key": "edition-drbench-iclr-2026-d00a3dcdb6",
"extent": {
"type": "json-value",
"value": "complete set of 100 tasks"
},
"id": "em:dossier-span:sha256:4aa9441f93251fa969f1fe8cc8a4fb6cf728ec3f8879efae3ffff55fabae759d",
"key": "span-c27aaca02bda06ed764be53351158fc862af9d4a556d3e82-104e7cda52",
"locator": {
"label": "page 5, Section 4.1",
"pointer": "/excerpts/2/text",
"type": "json-pointer"
},
"visibility": "public"
},
{
"digest": "sha256:2e8262448070ca7bcc1ea305d2d87a722f34721737d455d55f62185010bd2e32",
"edition_key": "edition-drbench-repo-main-469cce5-4f8964f3e3",
"extent": {
"type": "json-value",
"value": "100 PhD-level research tasks"
},
"id": "em:dossier-span:sha256:59d1cfd2232e3f962c838993d6a5456b6a1f66dfb989bcee79f4fc7b5b76a378",
"key": "span-0abbab3d6269d58eb76d8e90eff1e4c27512cf46d9b8e5ff-54e5759a8b",
"locator": {
"label": "repository README, Overview",
"pointer": "/excerpts/0/text",
"type": "json-pointer"
},
"visibility": "public"
},
{
"digest": "sha256:549018c6df261f27bd10133ad4e4f565e004f07710095259a7889cc79d149794",
"edition_key": "edition-drbench-repo-main-469cce5-4f8964f3e3",
"extent": {
"type": "json-value",
"value": "Support Verification: Uses web scraping and LLM judgment to verify whether cited sources actually support the claims"
},
"id": "em:dossier-span:sha256:a28651fed7ffef6fdbbe7f322faa2482aa1e0f6b1e2a0a909eddef59a72740f2",
"key": "span-92eb7cf19745e24190dda84c77d590941134d09f06805c0f-88afe43377",
"locator": {
"label": "README FACT, lines 240-248",
"pointer": "/excerpts/1/text",
"type": "json-pointer"
},
"visibility": "public"
},
{
"digest": "sha256:02cd1c04d6312fcda76549885095d9c9d971cc14709ae8f749cf47fc2c11bcee",
"edition_key": "edition-drbench-repo-main-469cce5-4f8964f3e3",
"extent": {
"type": "json-value",
"value": "Legacy code: the previous Gemini-2.5-Pro / Gemini-2.5-Flash evaluation code is preserved on the Gemini-2.5 branch."
},
"id": "em:dossier-span:sha256:abc4ab811e5c574575a0024557dda57a45d8d439ad4ff4ce08bef3753f7a86ef",
"key": "span-e5bcf0241a1aeae7d9a92755b1bcf0adeeea93a8922f22f8-77a5ff4cc0",
"locator": {
"label": "README News, 2026-05-11 evaluator migration notice",
"pointer": "/excerpts/2/text",
"type": "json-pointer"
},
"visibility": "public"
},
{
"digest": "sha256:9c713be4fc872b46e66d812e033708b51b4e385c81aa45f7b1798459a677343d",
"edition_key": "edition-drbench-repo-main-469cce5-4f8964f3e3",
"extent": {
"type": "json-value",
"value": "Official Evaluator Switched to GPT-5.5 ... Legacy code: the previous Gemini-2.5-Pro / Gemini-2.5-Flash evaluation code is preserved"
},
"id": "em:dossier-span:sha256:fd761963efd1f957ebe8dae1eb76239c2d17ce1553109a4e0ac420525c73cfcb",
"key": "span-ef637b38f36195bea09205d532e8744daf3e257f93d97144-26007673a3",
"locator": {
"label": "README News, lines 176-188",
"pointer": "/excerpts/3/text",
"type": "json-pointer"
},
"visibility": "public"
},
{
"digest": "sha256:e3ec5c81f615c657f7c6b124c5b2150d86f18281384ca0bd24ebb43a3317796f",
"edition_key": "edition-keplinger-supplement-v1-f6ddb4c04a",
"extent": {
"type": "json-value",
"value": "sentence-level evaluation"
},
"id": "em:dossier-span:sha256:38bf64d13c4edef9c0aba131c84d42e0e08637a661d2552cbf2cb49536813488",
"key": "span-94ad583514aa7e04b8d5f7b8f9cdda6fdffe8ca6712f033d-63c1ba24eb",
"locator": {
"label": "Dataset description",
"pointer": "/excerpts/0/text",
"type": "json-pointer"
},
"visibility": "public"
},
{
"digest": "sha256:c5ab0cef07fb26a5758cd8b25a249016665c446b524e1443a10c4c8929938d21",
"edition_key": "edition-keplinger-supplement-v2-01e44c09d8",
"extent": {
"type": "json-value",
"value": "Prompts for Deep Research-generated reviews and evaluation results"
},
"id": "em:dossier-span:sha256:5a849ff87effb5df1ca3dfa79cb62b7cc3c30c6a5197cef4a5330f49bf062dd2",
"key": "span-6f79c3c824ccf889d73e234335661e5fca68d90f8c78ee97-dbf7848a72",
"locator": {
"label": "dataset page, Description",
"pointer": "/excerpts/0/text",
"type": "json-pointer"
},
"visibility": "public"
},
{
"digest": "sha256:39f799b4bf1bcc50a74e5b019c585b970fa01779cf4b07bdeb27d71f35ea455b",
"edition_key": "edition-keplinger-vor-2025-10ba10f860",
"extent": {
"type": "json-value",
"value": "citation-bearing sentences exhibited high error rates: 51.3 ± 6.5% for ChatGPT and 57.8 ± 22.7% for Le Chat"
},
"id": "em:dossier-span:sha256:9f91508aa7be57f8e16c902e5399763ce349e7318b10d6085b65f0f53d5bdb56",
"key": "span-092745c7db6407d4d52927642741cde229fad05a5347bb18-7b04227f8f",
"locator": {
"label": "Main text, paragraph beginning 'For their high performances in generating reference lists'",
"pointer": "/excerpts/0/text",
"type": "json-pointer"
},
"visibility": "public"
},
{
"digest": "sha256:b6c3ff6f2ff60f37ae1c6b773ba6aa5a6c01015b61e3cc6f2a06bfcef69b53ad",
"edition_key": "edition-keplinger-vor-2025-10ba10f860",
"extent": {
"type": "json-value",
"value": "16 (69.6%)"
},
"id": "em:dossier-span:sha256:72a3edb3976aa397d38e9a7ccbc18b5e4af6f00c91540cd47aadcf9eddce2439",
"key": "span-39a9931f05ea356fe904ed15768fa08c66440b1fcb20f98e-191c24bbb2",
"locator": {
"label": "Table 1, ChatGPT all-correct subtotal",
"pointer": "/excerpts/1/text",
"type": "json-pointer"
},
"visibility": "public"
},
{
"digest": "sha256:39f799b4bf1bcc50a74e5b019c585b970fa01779cf4b07bdeb27d71f35ea455b",
"edition_key": "edition-keplinger-vor-2025-10ba10f860",
"extent": {
"type": "json-value",
"value": "citation-bearing sentences exhibited high error rates: 51.3 ± 6.5% for ChatGPT and 57.8 ± 22.7% for Le Chat"
},
"id": "em:dossier-span:sha256:910f7758be3883fc0964cd1e5412b9c75a586e6c4c7c08e64ebda5b9f541a132",
"key": "span-46c392bc942cd88d525d74298e6bc6ebd9aeaec533ddfb7b-774f5b996e",
"locator": {
"label": "main text, paragraph beginning 'For their high performances'; Figure 1",
"pointer": "/excerpts/2/text",
"type": "json-pointer"
},
"visibility": "public"
},
{
"digest": "sha256:5428bd1a04d5d2e332015107e915fe3a38f1d5e173e8c7246b743d98ecfd8618",
"edition_key": "edition-keplinger-vor-2025-10ba10f860",
"extent": {
"type": "json-value",
"value": "23 (100.0%)"
},
"id": "em:dossier-span:sha256:d9dab2e8e8da55f5a5ddbe3e62c658a7f643cd05c36c117aa8a0dfc76a7b6d09",
"key": "span-8c8d9a6dc003bf0298542b85df3f6f08db631e68e89f606f-92731379f2",
"locator": {
"label": "Table 1, ChatGPT subtotal",
"pointer": "/excerpts/3/text",
"type": "json-pointer"
},
"visibility": "public"
},
{
"digest": "sha256:1c94be957cfe750f445dc9143722905b5784b02f754a9b4071a3381e2bd493e9",
"edition_key": "edition-keplinger-vor-2025-10ba10f860",
"extent": {
"type": "json-value",
"value": "We imposed a 1500-word limit, APA-style references and conducted three independent runs."
},
"id": "em:dossier-span:sha256:54d8710861e7e1ee03c50697f1316c12138246e303a08c843fdadecfa99b8c5e",
"key": "span-b38a2af5982b4c85bc205d2f533a23ed3a8f40a49b651a5f-27395a6291",
"locator": {
"label": "Main text, paragraph beginning 'To address the gap' (search-indexed primary full text)",
"pointer": "/excerpts/4/text",
"type": "json-pointer"
},
"visibility": "public"
},
{
"digest": "sha256:28141fd1b0222658eb93a338783358926de770223eea24f8c5cbcf632d67f955",
"edition_key": "edition-keplinger-vor-2025-10ba10f860",
"extent": {
"type": "json-value",
"value": "ChatGPT Deep Research-generated references were mostly identifiable by their metadata (95.7%): 69.6% were entirely correct"
},
"id": "em:dossier-span:sha256:b9e4b58f255c32e2a9e5e6796e8d35f7f28d7481060d6b35f1e02a3c74d2b3d6",
"key": "span-b6a6ba48a2c4e352c8565b671a15a90400739f2aeefda8a6-e8232daf27",
"locator": {
"label": "Main text, following paragraph (search-indexed primary full text)",
"pointer": "/excerpts/5/text",
"type": "json-pointer"
},
"visibility": "public"
},
{
"digest": "sha256:2e43a0652065a6a1fb69de224082ac79d903cc71cac0664a1cb376cbfa09d0e8",
"edition_key": "edition-keplinger-vor-2025-10ba10f860",
"extent": {
"type": "json-value",
"value": "high error rates: 51.3 ± 6.5% for ChatGPT and 57.8 ± 22.7% for Le Chat"
},
"id": "em:dossier-span:sha256:f687ae7c3889980d4d3eea1b63b7bea25ae674ebc04caf8c59e7ae4f4fb77464",
"key": "span-b903dc2284e1b4d80bb6ef948d79bbdbc7d7447e871281b8-6ef6234316",
"locator": {
"label": "body, paragraph immediately after Table 1",
"pointer": "/excerpts/6/text",
"type": "json-pointer"
},
"visibility": "public"
},
{
"digest": "sha256:1f863c1f35c37af1fd2c3e4927ea0b289d94a9532fa5855e209ce98ea344d498",
"edition_key": "edition-keplinger-vor-2025-10ba10f860",
"extent": {
"type": "json-value",
"value": "51.3 ± 6.5%"
},
"id": "em:dossier-span:sha256:2f331f1e2636907b0cab1000b0947e440fa5f24eec0edaf9498470522e1abe1a",
"key": "span-f05b3189dc3fafd45b5bde643a91fd0616832928215a37a2-47f72e77e4",
"locator": {
"label": "Paragraph after Table 1",
"pointer": "/excerpts/7/text",
"type": "json-pointer"
},
"visibility": "public"
},
{
"digest": "sha256:064c8e109876030c9fdc887811e31b1e5505c4bb3968369b65b26566813cca3e",
"edition_key": "edition-liveresearchbench-arxiv-v5-f5644763e5",
"extent": {
"type": "json-value",
"value": "\"E1: URL is inaccessible or does not resolve; E2: URL content is irrelevant ... E3: URL content does not support the specific statements.\""
},
"id": "em:dossier-span:sha256:bebe5aa8ac392afd6d6bf674c11671ebdbfda99e8b197603f13e3d944bed7ea3",
"key": "span-d880991ad784390029d49f84949d083c4789d67d45893d87-c6ef8a5c13",
"locator": {
"label": "p.33, Appendix E",
"pointer": "/excerpts/0/text",
"type": "json-pointer"
},
"visibility": "public"
},
{
"digest": "sha256:0d570d54a0712207d83f71aae02fc0c8ca424a120d61b36c0d0e243611d8f842",
"edition_key": "edition-liveresearchbench-arxiv-v5-f5644763e5",
"extent": {
"type": "json-value",
"value": "\"human evaluators agree with either Gemini or GPT-5 on 87.1%\""
},
"id": "em:dossier-span:sha256:2d12624d0f0b1c18bd17cd81ce1742f1daf339a4d5dc188faf576502baa1e6ed",
"key": "span-e6f447129d9a662fb202d4ac22e16822bacfdbcbcdb7afe3-f7abe2194c",
"locator": {
"label": "p.23, Appendix C, Citation Accuracy",
"pointer": "/excerpts/1/text",
"type": "json-pointer"
},
"visibility": "public"
},
{
"digest": "sha256:ee6c65565e833960bd4ef6552fe3153a1108a743989a21b261b571cf6f066aa4",
"edition_key": "edition-reportbench-arxiv-v1-25888dc44b",
"extent": {
"type": "json-value",
"value": "\"most of which are concentrated in STEM fields\""
},
"id": "em:dossier-span:sha256:bc41155b5354576b75743d3e75b4e6242ea19e93d211c82faceeb28b3974c083",
"key": "span-084f11e57cf779850ea87aaf30ba1277438ff9efcb533fd2-fe72b27cfb",
"locator": {
"label": "Appendix A.1 'Limitations'",
"pointer": "/excerpts/0/text",
"type": "json-pointer"
},
"visibility": "public"
},
{
"digest": "sha256:4947edaf643bd0a234e0ca914d920a1a5b068d18b017dcfe6d6103c466d422ca",
"edition_key": "edition-reportbench-arxiv-v1-25888dc44b",
"extent": {
"type": "json-value",
"value": "The benchmark primarily draws from peer-reviewed survey papers on arXiv, most of which are concentrated in STEM fields."
},
"id": "em:dossier-span:sha256:bf94ce1ab912090b1db248bca0eae62fb77c0f4f12f2fffca9cafaf6524deadd",
"key": "span-1db12b68cfeb4c5f5a96fa9cd2e2bd46e8dfb4f096f73154-cea47f0c65",
"locator": {
"label": "Limitations, lines 259-261",
"pointer": "/excerpts/1/text",
"type": "json-pointer"
},
"visibility": "public"
},
{
"digest": "sha256:ac4714fdd34ead02999923441a133a798a67254f0449f293905ef5b1d1caffab",
"edition_key": "edition-reportbench-arxiv-v1-25888dc44b",
"extent": {
"type": "json-value",
"value": "the cited URL does not exist"
},
"id": "em:dossier-span:sha256:e896138fd05fe0904311d35606553601c77784112c448651ce822a85e31ed99c",
"key": "span-3015a109e7898cb720e426e9a57c526277b648c75f1fb4b1-745d84ff44",
"locator": {
"label": "Section 4, line 186",
"pointer": "/excerpts/2/text",
"type": "json-pointer"
},
"visibility": "public"
},
{
"digest": "sha256:4545fdd46cab2ffc151c9ddbfaab08b82862b68eebf5205cda180b5875a3aa81",
"edition_key": "edition-reportbench-arxiv-v1-25888dc44b",
"extent": {
"type": "json-value",
"value": "78.87% vs. 72.94%"
},
"id": "em:dossier-span:sha256:d87df9d6e418184717863f7c621ddc6c16838dc569a1ab2a558421fbc20eeaa9",
"key": "span-6c1d327e58ef001533ea1c11010e974436bbeafbde138efc-f6efd8d376",
"locator": {
"label": "Section 3.3, line 155",
"pointer": "/excerpts/3/text",
"type": "json-pointer"
},
"visibility": "public"
},
{
"digest": "sha256:e45fbf61feb38fc8de32433d439fdb8544133e905103b884c92327930a929b0a",
"edition_key": "edition-reportbench-arxiv-v1-25888dc44b",
"extent": {
"type": "json-value",
"value": "\"we manually collected responses ... during the period from July 14 to July 25... OpenAI was using the standard version of Deep Research, powered by the o3 model.\""
},
"id": "em:dossier-span:sha256:3c149d37c88677f94f1715b536ad1808c4e476392f7e02212adbc723e30d8e27",
"key": "span-7a6046a756d6ec0315d330cf0817d062ea48498bdc9c90b3-c3a1e80029",
"locator": {
"label": "p. 5, lines 315-329",
"pointer": "/excerpts/4/text",
"type": "json-pointer"
},
"visibility": "public"
},
{
"digest": "sha256:153b2f37a8662ac0dbbd7fe51bcbb291afd5d227c4da3f4c6463274c63a11fb2",
"edition_key": "edition-reportbench-arxiv-v1-25888dc44b",
"extent": {
"type": "json-value",
"value": "\"retrieve the full content of each cited webpage\""
},
"id": "em:dossier-span:sha256:5da64a3b244d7aa2f4baebee7a32be87f2f4255de92db8287d1b11a2a396932d",
"key": "span-9c7a4c186cc85f185aa293a7c5a46c08f9dbbc1a9e63c449-feff07a6e2",
"locator": {
"label": "Section 2.2 'Cited statements'",
"pointer": "/excerpts/5/text",
"type": "json-pointer"
},
"visibility": "public"
},
{
"digest": "sha256:397d217cf5022ad78aa1072c2e82bcf3197e5a87cf4ef96ea81a78e1ad39b384",
"edition_key": "edition-reportbench-arxiv-v1-25888dc44b",
"extent": {
"type": "json-value",
"value": "For statement extraction, supporting source extraction, and semantic consistency verification, we adopt gpt-4o."
},
"id": "em:dossier-span:sha256:f3bc53b3bd401949fea37755e5943e417ceb79e044727a73f7219cfc824ab859",
"key": "span-ba6b8a8b7d9121ebb05d594176ceed3d1e9dbc637876783a-d85147b6b0",
"locator": {
"label": "Settings and metrics, lines 132-139",
"pointer": "/excerpts/6/text",
"type": "json-pointer"
},
"visibility": "public"
},
{
"digest": "sha256:b55cc3d05dbba5d160b75515ce407128fe2e2cf8c2d679517e79309d6ed6ab40",
"edition_key": "edition-reportbench-arxiv-v1-25888dc44b",
"extent": {
"type": "json-value",
"value": "\"during the period from July 14 to July 25\""
},
"id": "em:dossier-span:sha256:abc076cdbec4977f668ce36fa6e5edcfebfd81c18c90a4f7913fbb0f8d5fe33a",
"key": "span-ddf4320a6d302c13d28b6225eb7e79ac15583349e1b449d4-f89bb46179",
"locator": {
"label": "Section 3.1 'Setttings'",
"pointer": "/excerpts/7/text",
"type": "json-pointer"
},
"visibility": "public"
},
{
"digest": "sha256:9b1bf034e823976ac267434e9c67ee7e4038da04f3d2582a0de7d6d729fc6916",
"edition_key": "edition-reportbench-arxiv-v1-25888dc44b",
"extent": {
"type": "json-value",
"value": "OpenAI Deep Research ... Match Rate 78.87%"
},
"id": "em:dossier-span:sha256:5ea28918cbd966b0fc14072426f2b7132bfe47f03d327e0b6eaee0629c3740ca",
"key": "span-e223640cb0ed12c87ca1d2406f3276f30a3b8d2017dd4dc1-3cfecd76bf",
"locator": {
"label": "Table 1, OpenAI Deep Research row and cited-statements Match Rate column",
"pointer": "/excerpts/8/text",
"type": "json-pointer"
},
"visibility": "public"
},
{
"digest": "sha256:12525ade29e741070aec9df092b69ee828e7f4333a4c41c331784cfab48bef56",
"edition_key": "edition-researcherbench-arxiv-v1-39a8b256d7",
"extent": {
"type": "json-value",
"value": "\"whether the extracted content supports the corresponding claim\""
},
"id": "em:dossier-span:sha256:96fb10c5924d80b1bf350f219f220a730a2da95d14e01a679c3355d7fd8fc706",
"key": "span-886b50d94e3facabc487342cd2af5157cf956622ecfc3686-e7bdad0935",
"locator": {
"label": "Section 4.2.1, Citation Support Verification and Score Computation",
"pointer": "/excerpts/0/text",
"type": "json-pointer"
},
"visibility": "public"
},
{
"digest": "sha256:1eb2dcb713304fb76409f4ea450d6b06ba05e95e02110aed33dad456ab3d8ce8",
"edition_key": "edition-researcherbench-arxiv-v1-39a8b256d7",
"extent": {
"type": "json-value",
"value": "\"All evaluations were conducted between March and April in 2025\""
},
"id": "em:dossier-span:sha256:7f4968754938f5e1c4aa96fe8d0c81aed24a12268e7fd6fd09ee3b3b7c9b2579",
"key": "span-ea9cb2e1b0078c0857b789d61d0a0d26801cf66891864705-2f19633d83",
"locator": {
"label": "Section 5.1 and Appendix E, Table 4",
"pointer": "/excerpts/1/text",
"type": "json-pointer"
},
"visibility": "public"
},
{
"digest": "sha256:12525ade29e741070aec9df092b69ee828e7f4333a4c41c331784cfab48bef56",
"edition_key": "edition-researcherbench-arxiv-v1-39a8b256d7",
"extent": {
"type": "json-value",
"value": "\"whether the extracted content supports the corresponding claim\""
},
"id": "em:dossier-span:sha256:5aba6568d3cbc2f788868b9ad96059aba175e6976e6989dab2736e0de95355b1",
"key": "span-f96ad5daa81153a3689f52f5777fedfd70e8a87d44e7a1c9-3da72b8748",
"locator": {
"label": "Section 4.2.1 'Citation Support Verification' and Section 4.2 'Score Computation', equations (2)-(3)",
"pointer": "/excerpts/2/text",
"type": "json-pointer"
},
"visibility": "public"
},
{
"digest": "sha256:ee1de7374c67e208aa5d1331c20a08c8c83418309f1bd768d380620bfdc27901",
"edition_key": "edition-url-health-arxiv-v1-85fb934b53",
"extent": {
"type": "json-value",
"value": "\"This study focuses on URL-based citation hallucinations\"; \"fabricated snippets ... and invented bibliographic entries ... require separate systematic study.\""
},
"id": "em:dossier-span:sha256:179697ec62048a6465fa50dda34ff54530e700430013a878201b9dadde405cc0",
"key": "span-55de6a6ca9414a252592e9f77544a9259533dd713ce1f12f-0826216ffc",
"locator": {
"label": "pp. 1-2, lines 64-95",
"pointer": "/excerpts/0/text",
"type": "json-pointer"
},
"visibility": "public"
},
{
"digest": "sha256:19f5203d42cda0ad9b817dc8b5e7abf1b449fc7fa9792a8cdcbf6ea8782349b1",
"edition_key": "edition-url-health-arxiv-v1-85fb934b53",
"extent": {
"type": "json-value",
"value": "\"our non-resolving rates are lower bounds\""
},
"id": "em:dossier-span:sha256:c9ea9babc817442f7cd3a5dc96392c22f987c46d4f8684b7db9942aac245abc0",
"key": "span-59d0da64f738e1d8745f2bff3663eb89e273f93fe501afe3-ba69fc58eb",
"locator": {
"label": "Section 2 definition discussion and Section 3.3",
"pointer": "/excerpts/1/text",
"type": "json-pointer"
},
"visibility": "public"
},
{
"digest": "sha256:3a14d6f574bed17e76ef8b560bbca0aa27b3a50918e49461433e34ec7770a4c9",
"edition_key": "edition-url-health-arxiv-v1-85fb934b53",
"extent": {
"type": "json-value",
"value": "no archived snapshot exists at any timestamp"
},
"id": "em:dossier-span:sha256:73d72803976f4bb5529880a7185545e8b0ce5581409b9f64ba6ace3ac07c4e2c",
"key": "span-a18e778fac3d49e81167f05e09fbc361e91af8c0c01b0fec-ca1b18e206",
"locator": {
"label": "page 3, Section 3.3",
"pointer": "/excerpts/2/text",
"type": "json-pointer"
},
"visibility": "public"
},
{
"digest": "sha256:7d1d86a3ca815ee38aafe98b3deb50c72abbf9f21f79336f8ab4843801b2dba5",
"edition_key": "edition-url-health-arxiv-v1-85fb934b53",
"extent": {
"type": "json-value",
"value": "the non-resolving rate drops significantly for all three models"
},
"id": "em:dossier-span:sha256:d90d09f34963ab45eebda00e5606202b7a929a8e000663d691d6536936d7a85d",
"key": "span-b68153666e4e12aee8dc887fef0a61fa5f44b5f680a0b7eb-769c2cc17d",
"locator": {
"label": "Section 5.1 Results",
"pointer": "/excerpts/3/text",
"type": "json-pointer"
},
"visibility": "public"
},
{
"digest": "sha256:b240bd208cb4c322f3c9c242d01e13da53b3fd2a64e41b8b0f26653032e0ee54",
"edition_key": "edition-url-health-arxiv-v1-85fb934b53",
"extent": {
"type": "json-value",
"value": "\"URLs returning 4xx or 5xx status codes, connection errors, or timeouts are classified as non-resolving\""
},
"id": "em:dossier-span:sha256:57bdc827daee408c0042ec10327966b2b898d525449bee7ecbce1e02330f7f5f",
"key": "span-bb62928b5e0cb4e373b2f6bfeb579d8564d05da5588f8b15-771310f4ef",
"locator": {
"label": "Section 3.3 'URL extraction and classification'",
"pointer": "/excerpts/4/text",
"type": "json-pointer"
},
"visibility": "public"
},
{
"digest": "sha256:9741316a5669b3b6e1e3022dfe4bfc23f53569a2839c46889423478a0b3c058f",
"edition_key": "edition-url-health-arxiv-v1-85fb934b53",
"extent": {
"type": "json-value",
"value": "DRBench (17) comprises 100 multilingual research queries (Chinese and English) covering finance, science, and technology, with pre-collected outputs from 23 models."
},
"id": "em:dossier-span:sha256:1adfd92244c349596074ef5bf78e7edf7297e8966c13f429410a0f5871d89e45",
"key": "span-supplement-url-health-drbench-precollected-outputs-44923dff13",
"locator": {
"label": "Section 3 Experimental setup; Section 3.1 Datasets; section#S3.SS1; paragraph p#S3.SS1.p1.1",
"pointer": "/review_supplements/0/extent",
"type": "json-pointer"
},
"visibility": "public"
},
{
"digest": "sha256:ab05cfa54280c980dabd52b0d336f1dc65a5761925a54dad2d5c7b1cbfb094cb",
"edition_key": "edition-em0026-audit-projection",
"extent": {
"type": "json-value",
"value": {
"answer_bytes": 40476,
"answer_path": "research/how-we-know/agent-citation-lineage/answers-v2/V2-SOL-01.json",
"answer_sha256": "17a7cf97b717d7b520026b9eb52a7da59ddb9b5f698c2aeadfab3f38c91008a0",
"prompt_sha256": "d321a9cec7b5fe419157c0623e18ff0020cb0080fbe6dd9ed4fd20a0b896f670",
"reported_model_identity": "unknown",
"requested_model_profile": "gpt-5.6-sol",
"retrieval_infrastructure": "unknown",
"run_id": "V2-SOL-01",
"status": "completed",
"trace_path": "research/how-we-know/agent-citation-lineage/traces-v2/V2-SOL-01.json"
}
},
"id": "em:dossier-span:sha256:ead238d33280cfd1dd252936d47fc1351950f8efb4105085d161ac7c1e51fa97",
"key": "span-audit-report-1",
"locator": {
"label": "Captured report V2-SOL-01",
"pointer": "/reports/0",
"type": "json-pointer"
},
"visibility": "public"
},
{
"digest": "sha256:bc1bb44771b6465a2b8393d63ae156007978f188e7560b00783eb1ed438acdb7",
"edition_key": "edition-em0026-audit-projection",
"extent": {
"type": "json-value",
"value": {
"answer_bytes": 32376,
"answer_path": "research/how-we-know/agent-citation-lineage/answers-v2/V2-SOL-02.json",
"answer_sha256": "f72975b2b98532116ca9d2c510e2343df505bfce359ebfef595a0a35b77cb810",
"prompt_sha256": "d321a9cec7b5fe419157c0623e18ff0020cb0080fbe6dd9ed4fd20a0b896f670",
"reported_model_identity": "unknown",
"requested_model_profile": "gpt-5.6-sol",
"retrieval_infrastructure": "unknown",
"run_id": "V2-SOL-02",
"status": "completed",
"trace_path": "research/how-we-know/agent-citation-lineage/traces-v2/V2-SOL-02.json"
}
},
"id": "em:dossier-span:sha256:f73bff2099b76bbc9f1afd31ed664329603886431f50689968527122a7cdf56c",
"key": "span-audit-report-2",
"locator": {
"label": "Captured report V2-SOL-02",
"pointer": "/reports/1",
"type": "json-pointer"
},
"visibility": "public"
},
{
"digest": "sha256:115b6d6686d6b70980a370cc8857d821dc4bbb125314d437de37b90eb122bf24",
"edition_key": "edition-em0026-audit-projection",
"extent": {
"type": "json-value",
"value": {
"answer_bytes": 38504,
"answer_path": "research/how-we-know/agent-citation-lineage/answers-v2/V2-SOL-03.json",
"answer_sha256": "782a60cc19d88217694cd672bac6f659f44fa0bd0f006f64a65ea77386d82dc7",
"prompt_sha256": "d321a9cec7b5fe419157c0623e18ff0020cb0080fbe6dd9ed4fd20a0b896f670",
"reported_model_identity": "unknown",
"requested_model_profile": "gpt-5.6-sol",
"retrieval_infrastructure": "unknown",
"run_id": "V2-SOL-03",
"status": "completed",
"trace_path": "research/how-we-know/agent-citation-lineage/traces-v2/V2-SOL-03.json"
}
},
"id": "em:dossier-span:sha256:ee1ce13bbdab686b6cc50461dd2ba7aa87bb19562a08da7e919faa0f7bcaa22d",
"key": "span-audit-report-3",
"locator": {
"label": "Captured report V2-SOL-03",
"pointer": "/reports/2",
"type": "json-pointer"
},
"visibility": "public"
},
{
"digest": "sha256:9b12680ff0b6ae58b223ae43c7466cea6766a8d3ae9d7853718f9fc7364456d9",
"edition_key": "edition-em0026-audit-projection",
"extent": {
"type": "json-value",
"value": {
"answer_bytes": 36500,
"answer_path": "research/how-we-know/agent-citation-lineage/answers-v2/V2-SOL-04.json",
"answer_sha256": "216552854069e32a8dc4986c57f59dc3b9b56d44f01d527167b3f311384a9261",
"prompt_sha256": "d321a9cec7b5fe419157c0623e18ff0020cb0080fbe6dd9ed4fd20a0b896f670",
"reported_model_identity": "unknown",
"requested_model_profile": "gpt-5.6-sol",
"retrieval_infrastructure": "credential-free public web search, open, click, and find operations; no logged-in browser session or account connector",
"run_id": "V2-SOL-04",
"status": "completed",
"trace_path": "research/how-we-know/agent-citation-lineage/traces-v2/V2-SOL-04.json"
}
},
"id": "em:dossier-span:sha256:47b229a3ae9d109248f1e08146d1ec39f610bb9c305b6e89a467470008767f2d",
"key": "span-audit-report-4",
"locator": {
"label": "Captured report V2-SOL-04",
"pointer": "/reports/3",
"type": "json-pointer"
},
"visibility": "public"
},
{
"digest": "sha256:24d9813e27bf472ffab087e4aaad435a366f57ff17ad760179e0927c6ef4bc25",
"edition_key": "edition-em0026-audit-projection",
"extent": {
"type": "json-value",
"value": {
"answer_bytes": 22598,
"answer_path": "research/how-we-know/agent-citation-lineage/answers-v2/V2-TERRA-01.json",
"answer_sha256": "e90c36b6f172fbe291d75ce1ea9451b1160515a2b7b4e0ead66aa1a76cc8b735",
"prompt_sha256": "d321a9cec7b5fe419157c0623e18ff0020cb0080fbe6dd9ed4fd20a0b896f670",
"reported_model_identity": "unknown",
"requested_model_profile": "gpt-5.6-terra",
"retrieval_infrastructure": "unknown",
"run_id": "V2-TERRA-01",
"status": "completed",
"trace_path": "research/how-we-know/agent-citation-lineage/traces-v2/V2-TERRA-01.json"
}
},
"id": "em:dossier-span:sha256:1536ede905a25868f4b4c310625e3b3d3635caa8671ffa6a11accc271c9008bf",
"key": "span-audit-report-5",
"locator": {
"label": "Captured report V2-TERRA-01",
"pointer": "/reports/4",
"type": "json-pointer"
},
"visibility": "public"
},
{
"digest": "sha256:c346c4cbf73a7c4371bba92f797bf763691748d22999be8d6d54d3092cb8eaf4",
"edition_key": "edition-em0026-audit-projection",
"extent": {
"type": "json-value",
"value": {
"answer_bytes": 20907,
"answer_path": "research/how-we-know/agent-citation-lineage/answers-v2/V2-TERRA-02.json",
"answer_sha256": "819d320162a09909a73df00a4e1ebbdd42e67794cb02e3f7c379f601e10a0192",
"prompt_sha256": "d321a9cec7b5fe419157c0623e18ff0020cb0080fbe6dd9ed4fd20a0b896f670",
"reported_model_identity": "unknown",
"requested_model_profile": "gpt-5.6-terra",
"retrieval_infrastructure": "unknown",
"run_id": "V2-TERRA-02",
"status": "completed",
"trace_path": "research/how-we-know/agent-citation-lineage/traces-v2/V2-TERRA-02.json"
}
},
"id": "em:dossier-span:sha256:b3cc1cbb4cf7e79d03b4f563fc8f77e77e14dce58514ea39746a30ee22074d57",
"key": "span-audit-report-6",
"locator": {
"label": "Captured report V2-TERRA-02",
"pointer": "/reports/5",
"type": "json-pointer"
},
"visibility": "public"
},
{
"digest": "sha256:925eee84e3fa929a32635f5e6d40db5435bc81fc05b0754d5a8bb0a6c841d8aa",
"edition_key": "edition-em0026-audit-projection",
"extent": {
"type": "json-value",
"value": {
"answer_bytes": 23183,
"answer_path": "research/how-we-know/agent-citation-lineage/answers-v2/V2-TERRA-03.json",
"answer_sha256": "e211dfe49066bb9f4ff9e59b91173d3b0f45b7b1e8ca5135c7435f3882d7e264",
"prompt_sha256": "d321a9cec7b5fe419157c0623e18ff0020cb0080fbe6dd9ed4fd20a0b896f670",
"reported_model_identity": "unknown",
"requested_model_profile": "gpt-5.6-terra",
"retrieval_infrastructure": "unknown",
"run_id": "V2-TERRA-03",
"status": "completed",
"trace_path": "research/how-we-know/agent-citation-lineage/traces-v2/V2-TERRA-03.json"
}
},
"id": "em:dossier-span:sha256:3ccb8c023b080e15fae593a51f18cc30308a296042e870a191bdad87a88d4aee",
"key": "span-audit-report-7",
"locator": {
"label": "Captured report V2-TERRA-03",
"pointer": "/reports/6",
"type": "json-pointer"
},
"visibility": "public"
},
{
"digest": "sha256:c9198194e5a51ebafd593b2d7e2386afb6ee76460f01bf2958f54df04981c528",
"edition_key": "edition-em0026-audit-projection",
"extent": {
"type": "json-value",
"value": {
"answer_bytes": 30189,
"answer_path": "research/how-we-know/agent-citation-lineage/answers-v2/V2-TERRA-04.json",
"answer_sha256": "8deffa91b6bb52d1de4ef4965d0087eca3d3a8a96d971a5e89753c1928cc4ef4",
"prompt_sha256": "d321a9cec7b5fe419157c0623e18ff0020cb0080fbe6dd9ed4fd20a0b896f670",
"reported_model_identity": "unknown",
"requested_model_profile": "gpt-5.6-terra",
"retrieval_infrastructure": "unknown",
"run_id": "V2-TERRA-04",
"status": "completed",
"trace_path": "research/how-we-know/agent-citation-lineage/traces-v2/V2-TERRA-04.json"
}
},
"id": "em:dossier-span:sha256:58db8115527aba46c1706f0fc3999e811f0352f5e855ad09e56c8688118d94b8",
"key": "span-audit-report-8",
"locator": {
"label": "Captured report V2-TERRA-04",
"pointer": "/reports/7",
"type": "json-pointer"
},
"visibility": "public"
},
{
"digest": "sha256:66dea042d4ff7ea895c2ac1f3d32f2e518f60e2b3733858207d31c5f0fd91315",
"edition_key": "edition-em0026-audit-projection",
"extent": {
"type": "json-value",
"value": {
"canonical_proposition": "In one five-system dermatology-review audit, reference identifiability and metadata correctness did not establish sentence-level claim-citation concordance.",
"data_root": "data:keplinger-three-run-five-system-audit",
"derivation_root": "derivation:keplinger-published-tables-and-figure",
"independent_review_status": "pending",
"method_root": "method:human-reference-and-sentence-concordance-audit",
"review_status": "candidate-supported-with-scope",
"warrant_id": "warrant:keplinger-metadata-versus-support",
"work_ids": [
"work:keplinger-dermatology-audit",
"work:keplinger-supplement"
]
}
},
"id": "em:dossier-span:sha256:4c31fb55b4cdd1822f9a06ab87c38ecd7d7abbf8590df51c7e695b2fc0b49940",
"key": "span-audit-candidate-1",
"locator": {
"label": "Candidate warrant warrant:keplinger-metadata-versus-support",
"pointer": "/candidate_warrants/0",
"type": "json-pointer"
},
"visibility": "public"
},
{
"digest": "sha256:17cf24491579ad90f058427047f6ddbf2ba8eeb195d41ae0d45ea9b894316645",
"edition_key": "edition-em0026-audit-projection",
"extent": {
"type": "json-value",
"value": {
"canonical_proposition": "DeepResearch Bench evaluated binary support for deduplicated statement-URL pairs; numeric values are edition-specific.",
"data_root": "data:deepresearch-bench-100-tasks-and-system-outputs",
"derivation_root": "derivation:deepresearch-bench-edition-specific-tables",
"independent_review_status": "pending",
"method_root": "method:fact-statement-url-support-evaluator",
"review_status": "candidate-supported-with-edition-boundary",
"warrant_id": "warrant:deepresearch-bench-fact",
"work_ids": [
"work:deepresearch-bench-paper",
"work:deepresearch-bench-repository"
]
}
},
"id": "em:dossier-span:sha256:dfc8ecf46c0995e4a1aa19839c6695eca56f7b84f60fb759f5dbb78a2993f776",
"key": "span-audit-candidate-2",
"locator": {
"label": "Candidate warrant warrant:deepresearch-bench-fact",
"pointer": "/candidate_warrants/1",
"type": "json-pointer"
},
"visibility": "public"
},
{
"digest": "sha256:b08eeec2df5d7a0a6ad35ca5ffebdc61b08e46229146ff404eee53976d795729",
"edition_key": "edition-em0026-audit-projection",
"extent": {
"type": "json-value",
"value": {
"canonical_proposition": "DeepTRACE measured statement-source support and found between-system variation; its Gemini value is internally inconsistent between table and prose.",
"data_root": "data:deeptrace-303-questions-2727-outputs",
"derivation_root": "derivation:deeptrace-tables-and-prose",
"independent_review_status": "pending",
"method_root": "method:deeptrace-statement-source-matrix",
"review_status": "candidate-supported-except-disputed-gemini-number",
"warrant_id": "warrant:deeptrace-support-variation",
"work_ids": [
"work:deeptrace"
]
}
},
"id": "em:dossier-span:sha256:24d59b5f6b46868a9a937c8a2cef20c7845bb288c7849eb0c38023996987d310",
"key": "span-audit-candidate-3",
"locator": {
"label": "Candidate warrant warrant:deeptrace-support-variation",
"pointer": "/candidate_warrants/2",
"type": "json-pointer"
},
"visibility": "public"
},
{
"digest": "sha256:45ba78edce01fdb5284f55cc1da77faa657670ae65b72476c556a72e91e43a62",
"edition_key": "edition-em0026-audit-projection",
"extent": {
"type": "json-value",
"value": {
"canonical_proposition": "Cited but Not Verified operationalized link access, topical relevance, and factual support separately and reported materially different rates.",
"data_root": "data:cnv-130-drbench-and-browsecomp-queries",
"derivation_root": "derivation:cnv-main-table",
"independent_review_status": "pending",
"method_root": "method:cnv-ast-access-relevance-fact-check",
"review_status": "candidate-supported-with-upstream-dependence",
"warrant_id": "warrant:cnv-link-relevance-support-gap",
"work_ids": [
"work:cited-not-verified"
]
}
},
"id": "em:dossier-span:sha256:5bd7c5980db79d9beb9545f42f9cd5c25d73bc1d3834bded3486e19950bf1d4e",
"key": "span-audit-candidate-4",
"locator": {
"label": "Candidate warrant warrant:cnv-link-relevance-support-gap",
"pointer": "/candidate_warrants/3",
"type": "json-pointer"
},
"visibility": "public"
},
{
"digest": "sha256:4a48e3a7f3fca8ad7158f19c6369279f15f3df7f20fa1f1694c2c3b4fc7355bc",
"edition_key": "edition-em0026-audit-projection",
"extent": {
"type": "json-value",
"value": {
"canonical_proposition": "Within the Cited but Not Verified harness, the 2-to-150-call ablation reduced reported Fact Check scores for two setups without establishing a general causal law about search depth.",
"data_root": "data:cnv-depth-ablation-two-model-setups",
"derivation_root": "derivation:cnv-depth-ablation-tables",
"independent_review_status": "pending",
"method_root": "method:cnv-ast-access-relevance-fact-check",
"review_status": "candidate-supported-as-within-harness-association",
"warrant_id": "warrant:cnv-search-depth-ablation",
"work_ids": [
"work:cited-not-verified"
]
}
},
"id": "em:dossier-span:sha256:aecbf27cd34ae509e4f493dc98f64193e9180a3f895c5ec8ca94460ae79c9322",
"key": "span-audit-candidate-5",
"locator": {
"label": "Candidate warrant warrant:cnv-search-depth-ablation",
"pointer": "/candidate_warrants/4",
"type": "json-pointer"
},
"visibility": "public"
},
{
"digest": "sha256:85f68e3ea34f66b257534dab515563196e5d081b56206a9cb3b7d95e59256139",
"edition_key": "edition-em0026-audit-projection",
"extent": {
"type": "json-value",
"value": {
"canonical_proposition": "The URL-health study measured HTTP resolution and Wayback presence, not semantic claim support, on outputs including reused DeepResearch Bench material.",
"data_root": "data:url-health-drbench-and-expertqa-urls",
"derivation_root": "derivation:url-health-observational-results",
"independent_review_status": "pending",
"method_root": "method:http-resolution-wayback-classification",
"review_status": "candidate-supported-with-semantic-exclusion",
"warrant_id": "warrant:url-health-resolution",
"work_ids": [
"work:url-health"
]
}
},
"id": "em:dossier-span:sha256:fa51c9c5d28d415b59802927038ab3b07ba183818a137cfc65d63f6392c2e742",
"key": "span-audit-candidate-6",
"locator": {
"label": "Candidate warrant warrant:url-health-resolution",
"pointer": "/candidate_warrants/5",
"type": "json-pointer"
},
"visibility": "public"
},
{
"digest": "sha256:204b3bcf88ec8e369c91f95f42ab367d360d39377a208098c594ce117745aef7",
"edition_key": "edition-em0026-audit-projection",
"extent": {
"type": "json-value",
"value": {
"canonical_proposition": "ReportBench compared cited statements with retrieved cited-page content and reported sub-100-percent semantic match rates in its bounded survey-task setting.",
"data_root": "data:reportbench-100-survey-tasks",
"derivation_root": "derivation:reportbench-table-1",
"independent_review_status": "pending",
"method_root": "method:reportbench-statement-source-semantic-match",
"review_status": "candidate-supported-with-judge-limitation",
"warrant_id": "warrant:reportbench-match-rate",
"work_ids": [
"work:reportbench"
]
}
},
"id": "em:dossier-span:sha256:59e20b99333a11e803d309e91ee7d4c1d003bbcb563819908fbca49847123051",
"key": "span-audit-candidate-7",
"locator": {
"label": "Candidate warrant warrant:reportbench-match-rate",
"pointer": "/candidate_warrants/6",
"type": "json-pointer"
},
"visibility": "public"
},
{
"digest": "sha256:da47a19a65e8c1a6816fe474c99558e32442ba728c9fe0046c1ee24d79ec33d4",
"edition_key": "edition-em0026-audit-projection",
"extent": {
"type": "json-value",
"value": "warrant:citation-verifier-calibration"
},
"id": "em:dossier-span:sha256:755999bc2f4f9858e910f7436ae3e011343fb8b295c7ac25ebe37124b9fa9a79",
"key": "span-audit-pending-1",
"locator": {
"label": "Pending warrant warrant:citation-verifier-calibration",
"pointer": "/pending_warrant_ids/0",
"type": "json-pointer"
},
"visibility": "public"
},
{
"digest": "sha256:07e84c48376d9fd67eb771f8afb587a7a6dfd3c79b39f6a75cf979484dd8ba3b",
"edition_key": "edition-em0026-audit-projection",
"extent": {
"type": "json-value",
"value": "warrant:liveresearchbench-e1-e2-e3"
},
"id": "em:dossier-span:sha256:25d5884da89ce905f6f847d89c5189011d125a985f742cd6ca5381adf98e24d1",
"key": "span-audit-pending-2",
"locator": {
"label": "Pending warrant warrant:liveresearchbench-e1-e2-e3",
"pointer": "/pending_warrant_ids/1",
"type": "json-pointer"
},
"visibility": "public"
},
{
"digest": "sha256:0deb7adf491a1372f1e476b53382ec21d96fd1cc8220805162b4db5448b6479b",
"edition_key": "edition-em0026-audit-projection",
"extent": {
"type": "json-value",
"value": "warrant:researcherbench-faithfulness-groundedness"
},
"id": "em:dossier-span:sha256:29708c22846ae1643541c2bfa5a0af26b63612a2120899e2b05e5aab5fa9f65f",
"key": "span-audit-pending-3",
"locator": {
"label": "Pending warrant warrant:researcherbench-faithfulness-groundedness",
"pointer": "/pending_warrant_ids/2",
"type": "json-pointer"
},
"visibility": "public"
},
{
"digest": "sha256:f63084450bd599189fafb74a53dee69e32b46e5dc0267645d8c3e7567097c33f",
"edition_key": "edition-em0026-audit-projection",
"extent": {
"type": "json-value",
"value": "warrant:url-health-correction-loop"
},
"id": "em:dossier-span:sha256:b5fba95dd09b0e35a7dc8861e676db6dc3bfbbecf8aef1b03e6124ac8f7cece4",
"key": "span-audit-pending-4",
"locator": {
"label": "Pending warrant warrant:url-health-correction-loop",
"pointer": "/pending_warrant_ids/3",
"type": "json-pointer"
},
"visibility": "public"
},
{
"digest": "sha256:d83cbf0b384a1082b6e85130f4393a66f2131b88bda60b7560b250948982f92b",
"edition_key": "edition-em0026-audit-projection",
"extent": {
"type": "json-value",
"value": {
"candidate_warrant_roots": 7,
"captured_reports": 8,
"citation_occurrences": 48,
"cited_urls": 30,
"exact_span_roots": 72,
"examined_edition_roots": 14,
"inaccessible_citations": 3,
"independently_confirmed_warrant_roots": 0,
"invalid_citations": 0,
"non_resolving_citations": 0,
"raw_claim_occurrences": 52,
"raw_span_occurrences": 127,
"resolving_url_roots": 27,
"source_work_roots": 11,
"unresolved_citations": 34,
"unsupported_or_force_raised_claims": 20
}
},
"id": "em:dossier-span:sha256:a7e533b2b8aa78289d24fc7234a530dfbf6aa9d3f4328d1d195bc7e2cbade8a8",
"key": "span-audit-counts",
"locator": {
"label": "Relation-derived packet counts",
"pointer": "/derived_counts",
"type": "json-pointer"
},
"visibility": "public"
},
{
"digest": "sha256:eb01a327b7f8f02a1744b36f49832915a24d1f4380f98c814d26f73db8db5840",
"edition_key": "edition-em0026-audit-projection",
"extent": {
"type": "json-value",
"value": {
"em0029_review_supplement_spans": 3,
"independently_rejected_claim_occurrences": 9,
"pending_warrant_roots": 4
}
},
"id": "em:dossier-span:sha256:f874fe9ff6f98c908f25b734b51da4df2ea473bb06e774970b860e7255ccde13",
"key": "span-audit-dispositions",
"locator": {
"label": "Relation-derived review disposition counts",
"pointer": "/disposition_counts",
"type": "json-pointer"
},
"visibility": "public"
},
{
"digest": "sha256:3187d8163b884569c89ff10ba36039614969488eaed19a9d52e9a3ff70148a20",
"edition_key": "edition-em0026-audit-projection",
"extent": {
"type": "json-value",
"value": [
{
"citation_occurrence_id": "V2-SOL-01:s1_keplinger",
"correction_ids": [],
"edition_id": "edition:keplinger-vor-2025",
"license": "CC BY-NC 4.0",
"license_treatment": "quote-minimal attributed spans; no full-text redistribution",
"raw_source_id": "s1_keplinger",
"raw_title": "Assessment of Deep Research for dermatology literature reviews: Deep concern over the hype",
"readback": {
"captured_bytes": 111993,
"captured_sha256": "abbaa9f22035230d13152a68aaeecd8e83e40867c9dae4acd1d547e36c866793",
"edition_id": "edition:keplinger-vor-2025",
"http_status": 200,
"media_type": "text/html; charset=utf-8",
"redirect_chain": [],
"requested_url": "https://pmc.ncbi.nlm.nih.gov/articles/PMC13109748/",
"resolved_url": "https://pmc.ncbi.nlm.nih.gov/articles/PMC13109748/",
"retrieval_status": "retrieved"
},
"requested_url": "https://pmc.ncbi.nlm.nih.gov/articles/PMC13109748/",
"resolution_status": "unresolved",
"run_id": "V2-SOL-01",
"source_work_id": "work:keplinger-dermatology-audit",
"span_occurrence_ids": [
"V2-SOL-01:s1_keplinger:sp1a",
"V2-SOL-01:s1_keplinger:sp1b",
"V2-SOL-01:s1_keplinger:sp1c"
]
},
{
"citation_occurrence_id": "V2-SOL-01:s1a_keplinger_data",
"correction_ids": [
"correction:mendeley-file-readback"
],
"edition_id": "edition:keplinger-supplement-v2",
"license": "CC BY 4.0",
"license_treatment": "metadata and quote-minimal landing-page spans; file API required authentication and was not used",
"raw_source_id": "s1a_keplinger_data",
"raw_title": "Supplementary materials of the article: Assessment of Deep Research for Dermatology Literature Reviews: Deep Concern Over the Hype",
"readback": {
"captured_bytes": 116951,
"captured_sha256": "0a2c4c6ed54dd7e2e0ca6cc07aa19b18fa1439e506c2310db778094282d2cebc",
"edition_id": "edition:keplinger-supplement-v2",
"http_status": 200,
"media_type": "text/html; charset=utf-8",
"redirect_chain": [],
"requested_url": "https://data.mendeley.com/datasets/3s73z9zf3c/2",
"resolved_url": "https://data.mendeley.com/datasets/3s73z9zf3c/2",
"retrieval_status": "retrieved"
},
"requested_url": "https://data.mendeley.com/datasets/3s73z9zf3c/2",
"resolution_status": "unresolved",
"run_id": "V2-SOL-01",
"source_work_id": "work:keplinger-supplement",
"span_occurrence_ids": [
"V2-SOL-01:s1a_keplinger_data:sp1d"
]
},
{
"citation_occurrence_id": "V2-SOL-01:s2_drbench",
"correction_ids": [],
"edition_id": "edition:drbench-iclr-2026",
"license": "unknown",
"license_treatment": "metadata, digest, and quote-minimal attributed spans only",
"raw_source_id": "s2_drbench",
"raw_title": "DeepResearch Bench: A Comprehensive Benchmark for Deep Research Agents",
"readback": {
"captured_bytes": 2665857,
"captured_sha256": "8f80ce247f7cc355bb6773f36037e01cc3ba2e2082c77381eaadf0d2b92b021c",
"edition_id": "edition:drbench-iclr-2026",
"http_status": 200,
"media_type": "application/pdf",
"redirect_chain": [],
"requested_url": "https://proceedings.iclr.cc/paper_files/paper/2026/file/465f22be10e07b301c6ed58f0472f704-Paper-Conference.pdf",
"resolved_url": "https://proceedings.iclr.cc/paper_files/paper/2026/file/465f22be10e07b301c6ed58f0472f704-Paper-Conference.pdf",
"retrieval_status": "retrieved"
},
"requested_url": "https://proceedings.iclr.cc/paper_files/paper/2026/file/465f22be10e07b301c6ed58f0472f704-Paper-Conference.pdf",
"resolution_status": "unresolved",
"run_id": "V2-SOL-01",
"source_work_id": "work:deepresearch-bench-paper",
"span_occurrence_ids": [
"V2-SOL-01:s2_drbench:sp2a",
"V2-SOL-01:s2_drbench:sp2b",
"V2-SOL-01:s2_drbench:sp2c",
"V2-SOL-01:s2_drbench:sp2d"
]
},
{
"citation_occurrence_id": "V2-SOL-01:s3_deeptrace",
"correction_ids": [],
"edition_id": "edition:deeptrace-iclr-2026",
"license": "unknown",
"license_treatment": "metadata, digest, and quote-minimal attributed spans only",
"raw_source_id": "s3_deeptrace",
"raw_title": "DeepTRACE: Auditing Deep Research AI Systems for Tracking Reliability Across Citations and Evidence",
"readback": {
"captured_bytes": 2982839,
"captured_sha256": "dea4981c1066d0240a005f603b3b14419c7e32beb5191b3df847ceb26af3d6b6",
"edition_id": "edition:deeptrace-iclr-2026",
"http_status": 200,
"media_type": "application/pdf",
"redirect_chain": [],
"requested_url": "https://proceedings.iclr.cc/paper_files/paper/2026/file/ad08767706825033b99122332293033d-Paper-Conference.pdf",
"resolved_url": "https://proceedings.iclr.cc/paper_files/paper/2026/file/ad08767706825033b99122332293033d-Paper-Conference.pdf",
"retrieval_status": "retrieved"
},
"requested_url": "https://proceedings.iclr.cc/paper_files/paper/2026/file/ad08767706825033b99122332293033d-Paper-Conference.pdf",
"resolution_status": "unresolved",
"run_id": "V2-SOL-01",
"source_work_id": "work:deeptrace",
"span_occurrence_ids": [
"V2-SOL-01:s3_deeptrace:sp3a",
"V2-SOL-01:s3_deeptrace:sp3b",
"V2-SOL-01:s3_deeptrace:sp3c",
"V2-SOL-01:s3_deeptrace:sp3d"
]
},
{
"citation_occurrence_id": "V2-SOL-01:s4_cited_not_verified",
"correction_ids": [],
"edition_id": "edition:cnv-arxiv-v1",
"license": "CC BY 4.0",
"license_treatment": "quote-minimal attributed spans",
"raw_source_id": "s4_cited_not_verified",
"raw_title": "Cited but Not Verified: Parsing and Evaluating Source Attribution in LLM Deep Research Agents",
"readback": {
"captured_bytes": 43848,
"captured_sha256": "d1b6d476b3e81460dbc8821757775a8fa589e0579b77167ef0f33d35cb819cb5",
"edition_id": "edition:cnv-arxiv-v1",
"http_status": 200,
"media_type": "text/html; charset=utf-8",
"redirect_chain": [],
"requested_url": "https://arxiv.org/abs/2605.06635",
"resolved_url": "https://arxiv.org/abs/2605.06635",
"retrieval_status": "retrieved"
},
"requested_url": "https://arxiv.org/abs/2605.06635",
"resolution_status": "unresolved",
"run_id": "V2-SOL-01",
"source_work_id": "work:cited-not-verified",
"span_occurrence_ids": [
"V2-SOL-01:s4_cited_not_verified:sp4a",
"V2-SOL-01:s4_cited_not_verified:sp4b",
"V2-SOL-01:s4_cited_not_verified:sp4c"
]
},
{
"citation_occurrence_id": "V2-SOL-01:s5_url_health",
"correction_ids": [],
"edition_id": "edition:url-health-arxiv-v1",
"license": "CC0 1.0",
"license_treatment": "quote-minimal attributed spans",
"raw_source_id": "s5_url_health",
"raw_title": "Detecting and Correcting Reference Hallucinations in Commercial LLMs and Deep Research Agents",
"readback": {
"captured_bytes": 42408,
"captured_sha256": "3976893e82f91cc9e7826c0d2d08c6caab40778134dbab09ebbfe1f6f63f5397",
"edition_id": "edition:url-health-arxiv-v1",
"http_status": 200,
"media_type": "text/html; charset=utf-8",
"redirect_chain": [],
"requested_url": "https://arxiv.org/abs/2604.03173",
"resolved_url": "https://arxiv.org/abs/2604.03173",
"retrieval_status": "retrieved"
},
"requested_url": "https://arxiv.org/abs/2604.03173",
"resolution_status": "unresolved",
"run_id": "V2-SOL-01",
"source_work_id": "work:url-health",
"span_occurrence_ids": [
"V2-SOL-01:s5_url_health:sp5a",
"V2-SOL-01:s5_url_health:sp5b",
"V2-SOL-01:s5_url_health:sp5c"
]
},
{
"citation_occurrence_id": "V2-SOL-02:S1",
"correction_ids": [],
"edition_id": "edition:url-health-arxiv-v1",
"license": "CC0 1.0",
"license_treatment": "quote-minimal attributed spans",
"raw_source_id": "S1",
"raw_title": "Detecting and Correcting Reference Hallucinations in Commercial LLMs and Deep Research Agents",
"readback": {
"captured_bytes": 302498,
"captured_sha256": "332e5b5cb4b0ee7065b1bbc30436dfdfeafca8130fb87e32b0020026d8bbafe1",
"edition_id": "edition:url-health-arxiv-v1",
"http_status": 200,
"media_type": "text/html; charset=utf-8",
"redirect_chain": [],
"requested_url": "https://arxiv.org/html/2604.03173v1",
"resolved_url": "https://arxiv.org/html/2604.03173v1",
"retrieval_status": "retrieved"
},
"requested_url": "https://arxiv.org/html/2604.03173v1",
"resolution_status": "unresolved",
"run_id": "V2-SOL-02",
"source_work_id": "work:url-health",
"span_occurrence_ids": [
"V2-SOL-02:S1:S1_SPAN_1",
"V2-SOL-02:S1:S1_SPAN_2"
]
},
{
"citation_occurrence_id": "V2-SOL-02:S3",
"correction_ids": [
"correction:mendeley-file-readback"
],
"edition_id": "edition:keplinger-supplement-v2",
"license": "CC BY 4.0",
"license_treatment": "metadata and quote-minimal landing-page spans; file API required authentication and was not used",
"raw_source_id": "S3",
"raw_title": "Supplementary materials of the article: Assessment of Deep Research for Dermatology Literature Reviews: Deep Concern Over the Hype",
"readback": {
"captured_bytes": 116951,
"captured_sha256": "0a2c4c6ed54dd7e2e0ca6cc07aa19b18fa1439e506c2310db778094282d2cebc",
"edition_id": "edition:keplinger-supplement-v2",
"http_status": 200,
"media_type": "text/html; charset=utf-8",
"redirect_chain": [],
"requested_url": "https://data.mendeley.com/datasets/3s73z9zf3c/2",
"resolved_url": "https://data.mendeley.com/datasets/3s73z9zf3c/2",
"retrieval_status": "retrieved"
},
"requested_url": "https://data.mendeley.com/datasets/3s73z9zf3c/2",
"resolution_status": "unresolved",
"run_id": "V2-SOL-02",
"source_work_id": "work:keplinger-supplement",
"span_occurrence_ids": [
"V2-SOL-02:S3:S3_SPAN_1",
"V2-SOL-02:S3:S3_SPAN_2"
]
},
{
"citation_occurrence_id": "V2-SOL-02:S4",
"correction_ids": [],
"edition_id": "edition:drbench-iclr-2026",
"license": "unknown",
"license_treatment": "metadata, digest, and quote-minimal attributed spans only",
"raw_source_id": "S4",
"raw_title": "DeepResearch Bench: A Comprehensive Benchmark for Deep Research Agents",
"readback": {
"captured_bytes": 12692,
"captured_sha256": "a2d811e0d24af89d9bd4b19cc9e8fbc474c67ef3827648550a6560eca2a5e332",
"edition_id": "edition:drbench-iclr-2026",
"http_status": 403,
"media_type": "text/html",
"redirect_chain": [],
"requested_url": "https://openreview.net/pdf?id=hQ0K2Hhq7H",
"resolved_url": "https://openreview.net/pdf?id=hQ0K2Hhq7H",
"retrieval_status": "inaccessible"
},
"requested_url": "https://openreview.net/pdf?id=hQ0K2Hhq7H",
"resolution_status": "unresolved",
"run_id": "V2-SOL-02",
"source_work_id": "work:deepresearch-bench-paper",
"span_occurrence_ids": [
"V2-SOL-02:S4:S4_SPAN_1",
"V2-SOL-02:S4:S4_SPAN_2"
]
},
{
"citation_occurrence_id": "V2-SOL-02:S5",
"correction_ids": [],
"edition_id": "edition:reportbench-arxiv-v1",
"license": "CC BY 4.0",
"license_treatment": "quote-minimal attributed spans",
"raw_source_id": "S5",
"raw_title": "ReportBench: Evaluating Deep Research Agents via Academic Survey Tasks",
"readback": {
"captured_bytes": 157386,
"captured_sha256": "055e568189a402dd510c7c84be60f59465d9a001808936bc25cb0026ac25d267",
"edition_id": "edition:reportbench-arxiv-v1",
"http_status": 200,
"media_type": "text/html; charset=utf-8",
"redirect_chain": [],
"requested_url": "https://arxiv.org/html/2508.15804v1",
"resolved_url": "https://arxiv.org/html/2508.15804v1",
"retrieval_status": "retrieved"
},
"requested_url": "https://arxiv.org/html/2508.15804v1",
"resolution_status": "unresolved",
"run_id": "V2-SOL-02",
"source_work_id": "work:reportbench",
"span_occurrence_ids": [
"V2-SOL-02:S5:S5_SPAN_1",
"V2-SOL-02:S5:S5_SPAN_2"
]
},
{
"citation_occurrence_id": "V2-SOL-02:S7",
"correction_ids": [],
"edition_id": "edition:cnv-arxiv-v1",
"license": "CC BY 4.0",
"license_treatment": "quote-minimal attributed spans",
"raw_source_id": "S7",
"raw_title": "Cited but Not Verified: Parsing and Evaluating Source Attribution in LLM Deep Research Agents",
"readback": {
"captured_bytes": 138385,
"captured_sha256": "7c5e3c33f3122b07d176e6679cd6762a95a6babaf4a353686f019f905b5526ca",
"edition_id": "edition:cnv-arxiv-v1",
"http_status": 200,
"media_type": "text/html; charset=utf-8",
"redirect_chain": [],
"requested_url": "https://arxiv.org/html/2605.06635v1",
"resolved_url": "https://arxiv.org/html/2605.06635v1",
"retrieval_status": "retrieved"
},
"requested_url": "https://arxiv.org/html/2605.06635v1",
"resolution_status": "unresolved",
"run_id": "V2-SOL-02",
"source_work_id": "work:cited-not-verified",
"span_occurrence_ids": [
"V2-SOL-02:S7:S7_SPAN_1",
"V2-SOL-02:S7:S7_SPAN_2"
]
},
{
"citation_occurrence_id": "V2-SOL-03:s1",
"correction_ids": [],
"edition_id": "edition:keplinger-vor-2025",
"license": "CC BY-NC 4.0",
"license_treatment": "quote-minimal attributed spans; no full-text redistribution",
"raw_source_id": "s1",
"raw_title": "Assessment of Deep Research for dermatology literature reviews: Deep concern over the hype",
"readback": {
"captured_bytes": 5498,
"captured_sha256": "d74c1891ae09a2b89121f2a60ac837c89761a8086759e45db214ca68c1905cea",
"edition_id": "edition:keplinger-vor-2025",
"http_status": 403,
"media_type": "text/html; charset=UTF-8",
"redirect_chain": [],
"requested_url": "https://onlinelibrary.wiley.com/doi/10.1111/jdv.70035",
"resolved_url": "https://onlinelibrary.wiley.com/doi/10.1111/jdv.70035",
"retrieval_status": "inaccessible"
},
"requested_url": "https://onlinelibrary.wiley.com/doi/10.1111/jdv.70035",
"resolution_status": "unresolved",
"run_id": "V2-SOL-03",
"source_work_id": "work:keplinger-dermatology-audit",
"span_occurrence_ids": [
"V2-SOL-03:s1:s1_span1"
]
},
{
"citation_occurrence_id": "V2-SOL-03:s3",
"correction_ids": [],
"edition_id": "edition:deeptrace-arxiv-v1",
"license": "arXiv non-exclusive distribution license",
"license_treatment": "metadata, digest, and quote-minimal attributed spans only",
"raw_source_id": "s3",
"raw_title": "DeepTRACE: Auditing Deep Research AI Systems for Tracking Reliability Across Citations and Evidence",
"readback": {
"captured_bytes": 364455,
"captured_sha256": "0c28edecaedd882584e985caac1e14f4a03dd59e939e17ae755a3c5e07ae42b0",
"edition_id": "edition:deeptrace-arxiv-v1",
"http_status": 200,
"media_type": "text/html; charset=utf-8",
"redirect_chain": [],
"requested_url": "https://arxiv.org/html/2509.04499",
"resolved_url": "https://arxiv.org/html/2509.04499",
"retrieval_status": "retrieved"
},
"requested_url": "https://arxiv.org/html/2509.04499",
"resolution_status": "unresolved",
"run_id": "V2-SOL-03",
"source_work_id": "work:deeptrace",
"span_occurrence_ids": [
"V2-SOL-03:s3:s3_span1"
]
},
{
"citation_occurrence_id": "V2-SOL-03:s4",
"correction_ids": [],
"edition_id": "edition:url-health-arxiv-v1",
"license": "CC0 1.0",
"license_treatment": "quote-minimal attributed spans",
"raw_source_id": "s4",
"raw_title": "Detecting and Correcting Reference Hallucinations in Commercial LLMs and Deep Research Agents",
"readback": {
"captured_bytes": 302498,
"captured_sha256": "332e5b5cb4b0ee7065b1bbc30436dfdfeafca8130fb87e32b0020026d8bbafe1",
"edition_id": "edition:url-health-arxiv-v1",
"http_status": 200,
"media_type": "text/html; charset=utf-8",
"redirect_chain": [],
"requested_url": "https://arxiv.org/html/2604.03173",
"resolved_url": "https://arxiv.org/html/2604.03173",
"retrieval_status": "retrieved"
},
"requested_url": "https://arxiv.org/html/2604.03173",
"resolution_status": "unresolved",
"run_id": "V2-SOL-03",
"source_work_id": "work:url-health",
"span_occurrence_ids": [
"V2-SOL-03:s4:s4_span1",
"V2-SOL-03:s4:s4_span2",
"V2-SOL-03:s4:s4_span3"
]
},
{
"citation_occurrence_id": "V2-SOL-04:S6",
"correction_ids": [
"correction:mendeley-file-readback"
],
"edition_id": "edition:keplinger-supplement-v1",
"license": "CC BY 4.0",
"license_treatment": "metadata and quote-minimal landing-page spans",
"raw_source_id": "S6",
"raw_title": "Supplementary materials of the article: Assessment of Deep Research for Dermatology Literature Reviews: Deep Concern Over the Hype",
"readback": {
"captured_bytes": 115612,
"captured_sha256": "da6308937337b2dfd3f97dc62f817e4a7ec362329ede0fce4c9bcdfd99ad93b1",
"edition_id": "edition:keplinger-supplement-v1",
"http_status": 200,
"media_type": "text/html; charset=utf-8",
"redirect_chain": [],
"requested_url": "https://data.mendeley.com/datasets/3s73z9zf3c/1",
"resolved_url": "https://data.mendeley.com/datasets/3s73z9zf3c/1",
"retrieval_status": "retrieved"
},
"requested_url": "https://data.mendeley.com/datasets/3s73z9zf3c/1",
"resolution_status": "unresolved",
"run_id": "V2-SOL-04",
"source_work_id": "work:keplinger-supplement",
"span_occurrence_ids": [
"V2-SOL-04:S6:S6-SP1"
]
},
{
"citation_occurrence_id": "V2-TERRA-01:s1_deepresearchbench",
"correction_ids": [],
"edition_id": "edition:drbench-arxiv-v1",
"license": "CC BY 4.0",
"license_treatment": "quote-minimal attributed spans",
"raw_source_id": "s1_deepresearchbench",
"raw_title": "DeepResearch Bench: A Comprehensive Benchmark for Deep Research Agents",
"readback": {
"captured_bytes": 3678399,
"captured_sha256": "8fbf30398f5e62f8839f0c9c8609bbb9e3cd0b57ae27d4bf33cb5db2007d1118",
"edition_id": "edition:drbench-arxiv-v1",
"http_status": 200,
"media_type": "application/pdf",
"redirect_chain": [],
"requested_url": "https://deepresearch-bench.github.io/static/papers/deepresearch-bench.pdf",
"resolved_url": "https://deepresearch-bench.github.io/static/papers/deepresearch-bench.pdf",
"retrieval_status": "retrieved"
},
"requested_url": "https://deepresearch-bench.github.io/static/papers/deepresearch-bench.pdf",
"resolution_status": "unresolved",
"run_id": "V2-TERRA-01",
"source_work_id": "work:deepresearch-bench-paper",
"span_occurrence_ids": [
"V2-TERRA-01:s1_deepresearchbench:s1_method",
"V2-TERRA-01:s1_deepresearchbench:s1_table",
"V2-TERRA-01:s1_deepresearchbench:s1_dates"
]
},
{
"citation_occurrence_id": "V2-TERRA-01:s2_reportbench",
"correction_ids": [],
"edition_id": "edition:reportbench-arxiv-v1",
"license": "CC BY 4.0",
"license_treatment": "quote-minimal attributed spans",
"raw_source_id": "s2_reportbench",
"raw_title": "ReportBench: Evaluating Deep Research Agents via Academic Survey Tasks",
"readback": {
"captured_bytes": 729607,
"captured_sha256": "90730ad75011d460305caf45a5be33b1b5eb4126f7e7efc1506f63beda1f91d3",
"edition_id": "edition:reportbench-arxiv-v1",
"http_status": 200,
"media_type": "application/pdf",
"redirect_chain": [],
"requested_url": "https://arxiv.org/pdf/2508.15804",
"resolved_url": "https://arxiv.org/pdf/2508.15804",
"retrieval_status": "retrieved"
},
"requested_url": "https://arxiv.org/pdf/2508.15804",
"resolution_status": "unresolved",
"run_id": "V2-TERRA-01",
"source_work_id": "work:reportbench",
"span_occurrence_ids": [
"V2-TERRA-01:s2_reportbench:s2_method",
"V2-TERRA-01:s2_reportbench:s2_metric_table",
"V2-TERRA-01:s2_reportbench:s2_collection"
]
},
{
"citation_occurrence_id": "V2-TERRA-01:s3_researcherbench",
"correction_ids": [],
"edition_id": "edition:researcherbench-arxiv-v1",
"license": "arXiv non-exclusive distribution license",
"license_treatment": "metadata, digest, and quote-minimal attributed spans only",
"raw_source_id": "s3_researcherbench",
"raw_title": "ResearcherBench: Evaluating Deep AI Research Systems on the Frontiers of Scientific Inquiry",
"readback": {
"captured_bytes": 1566914,
"captured_sha256": "1571435270d3e781ab8d5254913af00a7dcf0efcb99ec8e6d3db87a7b5c92f02",
"edition_id": "edition:researcherbench-arxiv-v1",
"http_status": 200,
"media_type": "application/pdf",
"redirect_chain": [],
"requested_url": "https://arxiv.org/pdf/2507.16280",
"resolved_url": "https://arxiv.org/pdf/2507.16280",
"retrieval_status": "retrieved"
},
"requested_url": "https://arxiv.org/pdf/2507.16280",
"resolution_status": "unresolved",
"run_id": "V2-TERRA-01",
"source_work_id": "work:researcherbench",
"span_occurrence_ids": [
"V2-TERRA-01:s3_researcherbench:s3_definition",
"V2-TERRA-01:s3_researcherbench:s3_models_time",
"V2-TERRA-01:s3_researcherbench:s3_table"
]
},
{
"citation_occurrence_id": "V2-TERRA-01:s4_cited_not_verified",
"correction_ids": [],
"edition_id": "edition:cnv-arxiv-v1",
"license": "CC BY 4.0",
"license_treatment": "quote-minimal attributed spans",
"raw_source_id": "s4_cited_not_verified",
"raw_title": "Cited but Not Verified: Parsing and Evaluating Source Attribution in LLM Deep Research Agents",
"readback": {
"captured_bytes": 1424589,
"captured_sha256": "db5b7b7e3d9ce9fc6d6713b60f65f0d81b07090ded2a28be7f541d1c290230b5",
"edition_id": "edition:cnv-arxiv-v1",
"http_status": 200,
"media_type": "application/pdf",
"redirect_chain": [],
"requested_url": "https://arxiv.org/pdf/2605.06635",
"resolved_url": "https://arxiv.org/pdf/2605.06635",
"retrieval_status": "retrieved"
},
"requested_url": "https://arxiv.org/pdf/2605.06635",
"resolution_status": "unresolved",
"run_id": "V2-TERRA-01",
"source_work_id": "work:cited-not-verified",
"span_occurrence_ids": [
"V2-TERRA-01:s4_cited_not_verified:s4_definition_table",
"V2-TERRA-01:s4_cited_not_verified:s4_depth"
]
},
{
"citation_occurrence_id": "V2-TERRA-01:s5_urlhealth",
"correction_ids": [],
"edition_id": "edition:url-health-arxiv-v1",
"license": "CC0 1.0",
"license_treatment": "quote-minimal attributed spans",
"raw_source_id": "s5_urlhealth",
"raw_title": "Detecting and Correcting Reference Hallucinations in Commercial LLMs and Deep Research Agents",
"readback": {
"captured_bytes": 416944,
"captured_sha256": "e8145a9f62f2a2bf0c5f3e2607160c56e010d02da965b5f65dbfa8f07ddde06c",
"edition_id": "edition:url-health-arxiv-v1",
"http_status": 200,
"media_type": "application/pdf",
"redirect_chain": [],
"requested_url": "https://arxiv.org/pdf/2604.03173",
"resolved_url": "https://arxiv.org/pdf/2604.03173",
"retrieval_status": "retrieved"
},
"requested_url": "https://arxiv.org/pdf/2604.03173",
"resolution_status": "unresolved",
"run_id": "V2-TERRA-01",
"source_work_id": "work:url-health",
"span_occurrence_ids": [
"V2-TERRA-01:s5_urlhealth:s5_scope",
"V2-TERRA-01:s5_urlhealth:s5_table",
"V2-TERRA-01:s5_urlhealth:s5_comparison"
]
},
{
"citation_occurrence_id": "V2-TERRA-02:s1_deepresearchbench",
"correction_ids": [],
"edition_id": "edition:drbench-arxiv-v1",
"license": "CC BY 4.0",
"license_treatment": "quote-minimal attributed spans",
"raw_source_id": "s1_deepresearchbench",
"raw_title": "DeepResearch Bench: A Comprehensive Benchmark for Deep Research Agents",
"readback": {
"captured_bytes": 3678399,
"captured_sha256": "8fbf30398f5e62f8839f0c9c8609bbb9e3cd0b57ae27d4bf33cb5db2007d1118",
"edition_id": "edition:drbench-arxiv-v1",
"http_status": 200,
"media_type": "application/pdf",
"redirect_chain": [],
"requested_url": "https://deepresearch-bench.github.io/static/papers/deepresearch-bench.pdf",
"resolved_url": "https://deepresearch-bench.github.io/static/papers/deepresearch-bench.pdf",
"retrieval_status": "retrieved"
},
"requested_url": "https://deepresearch-bench.github.io/static/papers/deepresearch-bench.pdf",
"resolution_status": "unresolved",
"run_id": "V2-TERRA-02",
"source_work_id": "work:deepresearch-bench-paper",
"span_occurrence_ids": [
"V2-TERRA-02:s1_deepresearchbench:s1_fact_method",
"V2-TERRA-02:s1_deepresearchbench:s1_fact_results",
"V2-TERRA-02:s1_deepresearchbench:s1_fact_formula",
"V2-TERRA-02:s1_deepresearchbench:s1_fact_judge_validation"
]
},
{
"citation_occurrence_id": "V2-TERRA-02:s2_liveresearchbench",
"correction_ids": [
"correction:liveresearchbench-license"
],
"edition_id": "edition:liveresearchbench-arxiv-v5",
"license": "CC BY-NC-SA 4.0",
"license_treatment": "quote-minimal attributed spans; raw CC BY claim is corrected in review records",
"raw_source_id": "s2_liveresearchbench",
"raw_title": "LiveResearchBench: A Live Benchmark for User-Centric Deep Research in the Wild",
"readback": {
"captured_bytes": 13057891,
"captured_sha256": "579b9728b76cfd242e9c94d9ff2985e196bbc72b5a741030e4f308ede04a4f69",
"edition_id": "edition:liveresearchbench-arxiv-v5",
"http_status": 200,
"media_type": "application/pdf",
"redirect_chain": [],
"requested_url": "https://arxiv.org/pdf/2510.14240v5",
"resolved_url": "https://arxiv.org/pdf/2510.14240v5",
"retrieval_status": "retrieved"
},
"requested_url": "https://arxiv.org/pdf/2510.14240v5",
"resolution_status": "unresolved",
"run_id": "V2-TERRA-02",
"source_work_id": "work:liveresearchbench",
"span_occurrence_ids": [
"V2-TERRA-02:s2_liveresearchbench:s2_rubric_tree",
"V2-TERRA-02:s2_liveresearchbench:s2_table7",
"V2-TERRA-02:s2_liveresearchbench:s2_validation"
]
},
{
"citation_occurrence_id": "V2-TERRA-02:s3_deeptrace",
"correction_ids": [],
"edition_id": "edition:deeptrace-iclr-2026",
"license": "unknown",
"license_treatment": "metadata, digest, and quote-minimal attributed spans only",
"raw_source_id": "s3_deeptrace",
"raw_title": "DeepTRACE: Auditing Deep Research AI Systems for Tracking Reliability Across Citations and Evidence",
"readback": {
"captured_bytes": 2982839,
"captured_sha256": "dea4981c1066d0240a005f603b3b14419c7e32beb5191b3df847ceb26af3d6b6",
"edition_id": "edition:deeptrace-iclr-2026",
"http_status": 200,
"media_type": "application/pdf",
"redirect_chain": [],
"requested_url": "https://proceedings.iclr.cc/paper_files/paper/2026/file/ad08767706825033b99122332293033d-Paper-Conference.pdf",
"resolved_url": "https://proceedings.iclr.cc/paper_files/paper/2026/file/ad08767706825033b99122332293033d-Paper-Conference.pdf",
"retrieval_status": "retrieved"
},
"requested_url": "https://proceedings.iclr.cc/paper_files/paper/2026/file/ad08767706825033b99122332293033d-Paper-Conference.pdf",
"resolution_status": "unresolved",
"run_id": "V2-TERRA-02",
"source_work_id": "work:deeptrace",
"span_occurrence_ids": [
"V2-TERRA-02:s3_deeptrace:s3_definition",
"V2-TERRA-02:s3_deeptrace:s3_table1",
"V2-TERRA-02:s3_deeptrace:s3_corpus_and_retrieval",
"V2-TERRA-02:s3_deeptrace:s3_judge_validation"
]
},
{
"citation_occurrence_id": "V2-TERRA-02:s4_researcherbench",
"correction_ids": [],
"edition_id": "edition:researcherbench-arxiv-v1",
"license": "arXiv non-exclusive distribution license",
"license_treatment": "metadata, digest, and quote-minimal attributed spans only",
"raw_source_id": "s4_researcherbench",
"raw_title": "ResearcherBench: Evaluating Deep AI Research Systems on the Frontiers of Scientific Inquiry",
"readback": {
"captured_bytes": 280402,
"captured_sha256": "c080d7304274d70a39651f49297cc65910ca833d7ee6a3de3052370733e44ee3",
"edition_id": "edition:researcherbench-arxiv-v1",
"http_status": 200,
"media_type": "text/html; charset=utf-8",
"redirect_chain": [],
"requested_url": "https://arxiv.org/html/2507.16280",
"resolved_url": "https://arxiv.org/html/2507.16280",
"retrieval_status": "retrieved"
},
"requested_url": "https://arxiv.org/html/2507.16280",
"resolution_status": "unresolved",
"run_id": "V2-TERRA-02",
"source_work_id": "work:researcherbench",
"span_occurrence_ids": [
"V2-TERRA-02:s4_researcherbench:s4_method",
"V2-TERRA-02:s4_researcherbench:s4_results"
]
},
{
"citation_occurrence_id": "V2-TERRA-03:S1",
"correction_ids": [],
"edition_id": "edition:cnv-arxiv-v1",
"license": "CC BY 4.0",
"license_treatment": "quote-minimal attributed spans",
"raw_source_id": "S1",
"raw_title": "Cited but Not Verified: Parsing and Evaluating Source Attribution in LLM Deep Research Agents",
"readback": {
"captured_bytes": 138385,
"captured_sha256": "7c5e3c33f3122b07d176e6679cd6762a95a6babaf4a353686f019f905b5526ca",
"edition_id": "edition:cnv-arxiv-v1",
"http_status": 200,
"media_type": "text/html; charset=utf-8",
"redirect_chain": [],
"requested_url": "https://arxiv.org/html/2605.06635",
"resolved_url": "https://arxiv.org/html/2605.06635",
"retrieval_status": "retrieved"
},
"requested_url": "https://arxiv.org/html/2605.06635",
"resolution_status": "unresolved",
"run_id": "V2-TERRA-03",
"source_work_id": "work:cited-not-verified",
"span_occurrence_ids": [
"V2-TERRA-03:S1:S1-A",
"V2-TERRA-03:S1:S1-B",
"V2-TERRA-03:S1:S1-C",
"V2-TERRA-03:S1:S1-D",
"V2-TERRA-03:S1:S1-E",
"V2-TERRA-03:S1:S1-F"
]
},
{
"citation_occurrence_id": "V2-TERRA-03:S2",
"correction_ids": [],
"edition_id": "edition:drbench-arxiv-v1",
"license": "CC BY 4.0",
"license_treatment": "quote-minimal attributed spans",
"raw_source_id": "S2",
"raw_title": "DeepResearch Bench: A Comprehensive Benchmark for Deep Research Agents",
"readback": {
"captured_bytes": 352002,
"captured_sha256": "9aa2894dbeaac30b23e7ffc8107a7f53b6e3855c8511838551de5bd7a422cc42",
"edition_id": "edition:drbench-arxiv-v1",
"http_status": 200,
"media_type": "text/html; charset=utf-8",
"redirect_chain": [],
"requested_url": "https://arxiv.org/html/2506.11763",
"resolved_url": "https://arxiv.org/html/2506.11763",
"retrieval_status": "retrieved"
},
"requested_url": "https://arxiv.org/html/2506.11763",
"resolution_status": "unresolved",
"run_id": "V2-TERRA-03",
"source_work_id": "work:deepresearch-bench-paper",
"span_occurrence_ids": [
"V2-TERRA-03:S2:S2-A",
"V2-TERRA-03:S2:S2-B",
"V2-TERRA-03:S2:S2-C",
"V2-TERRA-03:S2:S2-D"
]
},
{
"citation_occurrence_id": "V2-TERRA-03:S3",
"correction_ids": [],
"edition_id": "edition:reportbench-arxiv-v1",
"license": "CC BY 4.0",
"license_treatment": "quote-minimal attributed spans",
"raw_source_id": "S3",
"raw_title": "ReportBench: Evaluating Deep Research Agents via Academic Survey Tasks",
"readback": {
"captured_bytes": 157386,
"captured_sha256": "055e568189a402dd510c7c84be60f59465d9a001808936bc25cb0026ac25d267",
"edition_id": "edition:reportbench-arxiv-v1",
"http_status": 200,
"media_type": "text/html; charset=utf-8",
"redirect_chain": [],
"requested_url": "https://arxiv.org/html/2508.15804",
"resolved_url": "https://arxiv.org/html/2508.15804",
"retrieval_status": "retrieved"
},
"requested_url": "https://arxiv.org/html/2508.15804",
"resolution_status": "unresolved",
"run_id": "V2-TERRA-03",
"source_work_id": "work:reportbench",
"span_occurrence_ids": [
"V2-TERRA-03:S3:S3-A",
"V2-TERRA-03:S3:S3-B",
"V2-TERRA-03:S3:S3-C",
"V2-TERRA-03:S3:S3-D"
]
},
{
"citation_occurrence_id": "V2-TERRA-03:S4",
"correction_ids": [],
"edition_id": "edition:keplinger-vor-2025",
"license": "CC BY-NC 4.0",
"license_treatment": "quote-minimal attributed spans; no full-text redistribution",
"raw_source_id": "S4",
"raw_title": "Assessment of Deep Research for dermatology literature reviews: Deep concern over the hype",
"readback": {
"captured_bytes": 111993,
"captured_sha256": "abbaa9f22035230d13152a68aaeecd8e83e40867c9dae4acd1d547e36c866793",
"edition_id": "edition:keplinger-vor-2025",
"http_status": 200,
"media_type": "text/html; charset=utf-8",
"redirect_chain": [],
"requested_url": "https://pmc.ncbi.nlm.nih.gov/articles/PMC13109748/",
"resolved_url": "https://pmc.ncbi.nlm.nih.gov/articles/PMC13109748/",
"retrieval_status": "retrieved"
},
"requested_url": "https://pmc.ncbi.nlm.nih.gov/articles/PMC13109748/",
"resolution_status": "unresolved",
"run_id": "V2-TERRA-03",
"source_work_id": "work:keplinger-dermatology-audit",
"span_occurrence_ids": [
"V2-TERRA-03:S4:S4-A",
"V2-TERRA-03:S4:S4-B",
"V2-TERRA-03:S4:S4-C"
]
},
{
"citation_occurrence_id": "V2-TERRA-03:S6",
"correction_ids": [],
"edition_id": "edition:keplinger-vor-2025",
"license": "CC BY-NC 4.0",
"license_treatment": "quote-minimal attributed spans; no full-text redistribution",
"raw_source_id": "S6",
"raw_title": "PubMed record and Figure 1 caption for Assessment of Deep Research for dermatology literature reviews: Deep concern over the hype",
"readback": {
"captured_bytes": 5565,
"captured_sha256": "a46109544fe4ff4504fab5e97abea3cb7172367aba54a308937386394f0ff046",
"edition_id": "edition:keplinger-vor-2025",
"http_status": 203,
"media_type": "text/html; charset=utf-8",
"redirect_chain": [],
"requested_url": "https://pubmed.ncbi.nlm.nih.gov/40904191/",
"resolved_url": "https://pubmed.ncbi.nlm.nih.gov/40904191/",
"retrieval_status": "inaccessible"
},
"requested_url": "https://pubmed.ncbi.nlm.nih.gov/40904191/",
"resolution_status": "unresolved",
"run_id": "V2-TERRA-03",
"source_work_id": "work:keplinger-dermatology-audit",
"span_occurrence_ids": [
"V2-TERRA-03:S6:S6-A"
]
},
{
"citation_occurrence_id": "V2-TERRA-04:S1_deepresearchbench",
"correction_ids": [],
"edition_id": "edition:drbench-arxiv-v1",
"license": "CC BY 4.0",
"license_treatment": "quote-minimal attributed spans",
"raw_source_id": "S1_deepresearchbench",
"raw_title": "DeepResearch Bench: A Comprehensive Benchmark for Deep Research Agents",
"readback": {
"captured_bytes": 352002,
"captured_sha256": "9aa2894dbeaac30b23e7ffc8107a7f53b6e3855c8511838551de5bd7a422cc42",
"edition_id": "edition:drbench-arxiv-v1",
"http_status": 200,
"media_type": "text/html; charset=utf-8",
"redirect_chain": [],
"requested_url": "https://arxiv.org/html/2506.11763",
"resolved_url": "https://arxiv.org/html/2506.11763",
"retrieval_status": "retrieved"
},
"requested_url": "https://arxiv.org/html/2506.11763",
"resolution_status": "unresolved",
"run_id": "V2-TERRA-04",
"source_work_id": "work:deepresearch-bench-paper",
"span_occurrence_ids": [
"V2-TERRA-04:S1_deepresearchbench:S1_method",
"V2-TERRA-04:S1_deepresearchbench:S1_table1",
"V2-TERRA-04:S1_deepresearchbench:S1_metric",
"V2-TERRA-04:S1_deepresearchbench:S1_time"
]
},
{
"citation_occurrence_id": "V2-TERRA-04:S2_researcherbench",
"correction_ids": [],
"edition_id": "edition:researcherbench-arxiv-v1",
"license": "arXiv non-exclusive distribution license",
"license_treatment": "metadata, digest, and quote-minimal attributed spans only",
"raw_source_id": "S2_researcherbench",
"raw_title": "ResearcherBench: Evaluating Deep AI Research Systems on the Frontiers of Scientific Inquiry",
"readback": {
"captured_bytes": 280402,
"captured_sha256": "c080d7304274d70a39651f49297cc65910ca833d7ee6a3de3052370733e44ee3",
"edition_id": "edition:researcherbench-arxiv-v1",
"http_status": 200,
"media_type": "text/html; charset=utf-8",
"redirect_chain": [],
"requested_url": "https://arxiv.org/html/2507.16280",
"resolved_url": "https://arxiv.org/html/2507.16280",
"retrieval_status": "retrieved"
},
"requested_url": "https://arxiv.org/html/2507.16280",
"resolution_status": "unresolved",
"run_id": "V2-TERRA-04",
"source_work_id": "work:researcherbench",
"span_occurrence_ids": [
"V2-TERRA-04:S2_researcherbench:S2_method",
"V2-TERRA-04:S2_researcherbench:S2_table2",
"V2-TERRA-04:S2_researcherbench:S2_scope",
"V2-TERRA-04:S2_researcherbench:S2_time"
]
},
{
"citation_occurrence_id": "V2-TERRA-04:S3_reportbench",
"correction_ids": [],
"edition_id": "edition:reportbench-arxiv-v1",
"license": "CC BY 4.0",
"license_treatment": "quote-minimal attributed spans",
"raw_source_id": "S3_reportbench",
"raw_title": "ReportBench: Evaluating Deep Research Agents via Academic Survey Tasks",
"readback": {
"captured_bytes": 157386,
"captured_sha256": "055e568189a402dd510c7c84be60f59465d9a001808936bc25cb0026ac25d267",
"edition_id": "edition:reportbench-arxiv-v1",
"http_status": 200,
"media_type": "text/html; charset=utf-8",
"redirect_chain": [],
"requested_url": "https://arxiv.org/html/2508.15804",
"resolved_url": "https://arxiv.org/html/2508.15804",
"retrieval_status": "retrieved"
},
"requested_url": "https://arxiv.org/html/2508.15804",
"resolution_status": "unresolved",
"run_id": "V2-TERRA-04",
"source_work_id": "work:reportbench",
"span_occurrence_ids": [
"V2-TERRA-04:S3_reportbench:S3_method",
"V2-TERRA-04:S3_reportbench:S3_table1",
"V2-TERRA-04:S3_reportbench:S3_time",
"V2-TERRA-04:S3_reportbench:S3_limitations"
]
},
{
"citation_occurrence_id": "V2-TERRA-04:S4_urlhealth",
"correction_ids": [],
"edition_id": "edition:url-health-arxiv-v1",
"license": "CC0 1.0",
"license_treatment": "quote-minimal attributed spans",
"raw_source_id": "S4_urlhealth",
"raw_title": "Detecting and Correcting Reference Hallucinations in Commercial LLMs and Deep Research Agents",
"readback": {
"captured_bytes": 302498,
"captured_sha256": "332e5b5cb4b0ee7065b1bbc30436dfdfeafca8130fb87e32b0020026d8bbafe1",
"edition_id": "edition:url-health-arxiv-v1",
"http_status": 200,
"media_type": "text/html; charset=utf-8",
"redirect_chain": [],
"requested_url": "https://arxiv.org/html/2604.03173",
"resolved_url": "https://arxiv.org/html/2604.03173",
"retrieval_status": "retrieved"
},
"requested_url": "https://arxiv.org/html/2604.03173",
"resolution_status": "unresolved",
"run_id": "V2-TERRA-04",
"source_work_id": "work:url-health",
"span_occurrence_ids": [
"V2-TERRA-04:S4_urlhealth:S4_table2",
"V2-TERRA-04:S4_urlhealth:S4_comparison",
"V2-TERRA-04:S4_urlhealth:S4_method",
"V2-TERRA-04:S4_urlhealth:S4_limitations"
]
},
{
"citation_occurrence_id": "V2-TERRA-04:S5_cited_not_verified",
"correction_ids": [],
"edition_id": "edition:cnv-arxiv-v1",
"license": "CC BY 4.0",
"license_treatment": "quote-minimal attributed spans",
"raw_source_id": "S5_cited_not_verified",
"raw_title": "Cited but Not Verified: Parsing and Evaluating Source Attribution in LLM Deep Research Agents",
"readback": {
"captured_bytes": 138385,
"captured_sha256": "7c5e3c33f3122b07d176e6679cd6762a95a6babaf4a353686f019f905b5526ca",
"edition_id": "edition:cnv-arxiv-v1",
"http_status": 200,
"media_type": "text/html; charset=utf-8",
"redirect_chain": [],
"requested_url": "https://arxiv.org/html/2605.06635",
"resolved_url": "https://arxiv.org/html/2605.06635",
"retrieval_status": "retrieved"
},
"requested_url": "https://arxiv.org/html/2605.06635",
"resolution_status": "unresolved",
"run_id": "V2-TERRA-04",
"source_work_id": "work:cited-not-verified",
"span_occurrence_ids": [
"V2-TERRA-04:S5_cited_not_verified:S5_method",
"V2-TERRA-04:S5_cited_not_verified:S5_scope",
"V2-TERRA-04:S5_cited_not_verified:S5_table1",
"V2-TERRA-04:S5_cited_not_verified:S5_depth",
"V2-TERRA-04:S5_cited_not_verified:S5_limitations"
]
}
]
},
"id": "em:dossier-span:sha256:a399dcbcdf2310b4abfea59748d83a462942013aac90f34625abbff7c4f74a2e",
"key": "span-audit-unresolved",
"locator": {
"label": "All unresolved citation occurrences",
"pointer": "/unresolved_citations",
"type": "json-pointer"
},
"visibility": "public"
},
{
"digest": "sha256:98ee18b4c442784a4c37ac10bee772994b5e19978343aa8267197d4ab90028e6",
"edition_key": "edition-em0026-audit-projection",
"extent": {
"type": "json-value",
"value": [
{
"citation_occurrence_id": "V2-SOL-02:S4",
"correction_ids": [],
"edition_id": "edition:drbench-iclr-2026",
"license": "unknown",
"license_treatment": "metadata, digest, and quote-minimal attributed spans only",
"raw_source_id": "S4",
"raw_title": "DeepResearch Bench: A Comprehensive Benchmark for Deep Research Agents",
"readback": {
"captured_bytes": 12692,
"captured_sha256": "a2d811e0d24af89d9bd4b19cc9e8fbc474c67ef3827648550a6560eca2a5e332",
"edition_id": "edition:drbench-iclr-2026",
"http_status": 403,
"media_type": "text/html",
"redirect_chain": [],
"requested_url": "https://openreview.net/pdf?id=hQ0K2Hhq7H",
"resolved_url": "https://openreview.net/pdf?id=hQ0K2Hhq7H",
"retrieval_status": "inaccessible"
},
"requested_url": "https://openreview.net/pdf?id=hQ0K2Hhq7H",
"resolution_status": "unresolved",
"run_id": "V2-SOL-02",
"source_work_id": "work:deepresearch-bench-paper",
"span_occurrence_ids": [
"V2-SOL-02:S4:S4_SPAN_1",
"V2-SOL-02:S4:S4_SPAN_2"
]
},
{
"citation_occurrence_id": "V2-SOL-03:s1",
"correction_ids": [],
"edition_id": "edition:keplinger-vor-2025",
"license": "CC BY-NC 4.0",
"license_treatment": "quote-minimal attributed spans; no full-text redistribution",
"raw_source_id": "s1",
"raw_title": "Assessment of Deep Research for dermatology literature reviews: Deep concern over the hype",
"readback": {
"captured_bytes": 5498,
"captured_sha256": "d74c1891ae09a2b89121f2a60ac837c89761a8086759e45db214ca68c1905cea",
"edition_id": "edition:keplinger-vor-2025",
"http_status": 403,
"media_type": "text/html; charset=UTF-8",
"redirect_chain": [],
"requested_url": "https://onlinelibrary.wiley.com/doi/10.1111/jdv.70035",
"resolved_url": "https://onlinelibrary.wiley.com/doi/10.1111/jdv.70035",
"retrieval_status": "inaccessible"
},
"requested_url": "https://onlinelibrary.wiley.com/doi/10.1111/jdv.70035",
"resolution_status": "unresolved",
"run_id": "V2-SOL-03",
"source_work_id": "work:keplinger-dermatology-audit",
"span_occurrence_ids": [
"V2-SOL-03:s1:s1_span1"
]
},
{
"citation_occurrence_id": "V2-TERRA-03:S6",
"correction_ids": [],
"edition_id": "edition:keplinger-vor-2025",
"license": "CC BY-NC 4.0",
"license_treatment": "quote-minimal attributed spans; no full-text redistribution",
"raw_source_id": "S6",
"raw_title": "PubMed record and Figure 1 caption for Assessment of Deep Research for dermatology literature reviews: Deep concern over the hype",
"readback": {
"captured_bytes": 5565,
"captured_sha256": "a46109544fe4ff4504fab5e97abea3cb7172367aba54a308937386394f0ff046",
"edition_id": "edition:keplinger-vor-2025",
"http_status": 203,
"media_type": "text/html; charset=utf-8",
"redirect_chain": [],
"requested_url": "https://pubmed.ncbi.nlm.nih.gov/40904191/",
"resolved_url": "https://pubmed.ncbi.nlm.nih.gov/40904191/",
"retrieval_status": "inaccessible"
},
"requested_url": "https://pubmed.ncbi.nlm.nih.gov/40904191/",
"resolution_status": "unresolved",
"run_id": "V2-TERRA-03",
"source_work_id": "work:keplinger-dermatology-audit",
"span_occurrence_ids": [
"V2-TERRA-03:S6:S6-A"
]
}
]
},
"id": "em:dossier-span:sha256:fe4b10fe8b860e86a721cce5a2c2183c97c7fbdb3b69d7e5d9067a9c59783ae6",
"key": "span-audit-inaccessible",
"locator": {
"label": "All inaccessible citation carriers",
"pointer": "/inaccessible_citations",
"type": "json-pointer"
},
"visibility": "public"
},
{
"digest": "sha256:608e58805bcff58922f73ed2e0459b3cfcf1f8df9ee70e3d1238f68ac79712be",
"edition_key": "edition-em0026-audit-projection",
"extent": {
"type": "json-value",
"value": [
"V2-SOL-01:r3_deepresearch_bench_fact",
"V2-SOL-01:r4_deeptrace_support",
"V2-SOL-02:R3_deepresearch_bench_fact",
"V2-SOL-02:R5_deeptrace_audit",
"V2-SOL-03:r3_deepresearch_bench_fact",
"V2-SOL-03:r4_deeptrace_support",
"V2-SOL-03:r8_search_depth_ablation",
"V2-SOL-03:r9_verifier_calibration",
"V2-SOL-04:R1",
"V2-SOL-04:R2",
"V2-SOL-04:R3",
"V2-SOL-04:R4",
"V2-SOL-04:R5",
"V2-SOL-04:R8",
"V2-TERRA-01:r1_deepresearchbench_fact",
"V2-TERRA-01:r4_cited_not_verified_source_attribution",
"V2-TERRA-02:answer",
"V2-TERRA-02:r1_deepresearchbench_fact",
"V2-TERRA-02:r4_deeptrace",
"V2-TERRA-03:R3"
]
},
"id": "em:dossier-span:sha256:087792c927d013e13f7eed7aafd137b4b759473b95866ccdd1fbf5ffb662e65f",
"key": "span-audit-unsupported",
"locator": {
"label": "Unsupported or force-raised claim occurrences",
"pointer": "/unsupported_or_force_raised_occurrence_ids",
"type": "json-pointer"
},
"visibility": "public"
},
{
"digest": "sha256:fc0a27fd59473c91e1a40670b6c9c07d037b5d6a56a2e977b7e7b8be5e240236",
"edition_key": "edition-em0026-audit-projection",
"extent": {
"type": "json-value",
"value": [
"V2-SOL-02:R5_deeptrace_audit",
"V2-SOL-03:r3_deepresearch_bench_fact",
"V2-SOL-03:r8_search_depth_ablation",
"V2-SOL-03:r9_verifier_calibration",
"V2-SOL-04:R1",
"V2-SOL-04:R2",
"V2-SOL-04:R3",
"V2-SOL-04:R4",
"V2-SOL-04:R8"
]
},
"id": "em:dossier-span:sha256:54fe13d731b79401537cd8dcc38d0c59a13c51a68916545018521cf600d772b7",
"key": "span-audit-rejected",
"locator": {
"label": "Independently rejected claim occurrences",
"pointer": "/rejected_claim_occurrence_ids",
"type": "json-pointer"
},
"visibility": "public"
},
{
"digest": "sha256:92b558b23559a96a551de8acb2627b148688d21ede2ca3fb9966196c4675dab6",
"edition_key": "edition-em0026-audit-projection",
"extent": {
"type": "json-value",
"value": {
"automatic_independence_credit": 0,
"prompt_sha256": "d321a9cec7b5fe419157c0623e18ff0020cb0080fbe6dd9ed4fd20a0b896f670",
"reported_model_identity": "unknown",
"requested_profiles": [
"gpt-5.6-sol",
"gpt-5.6-terra"
],
"retrieval_infrastructure": "unknown"
}
},
"id": "em:dossier-span:sha256:e120064fee0bdabe15299cb865ff2f7df66473291a13b7a765a4a09d16223f5e",
"key": "span-audit-capture-lineage",
"locator": {
"label": "Shared capture lineage",
"pointer": "/shared_capture_lineage",
"type": "json-pointer"
},
"visibility": "public"
},
{
"digest": "sha256:3dfd932c8b71669ba1f28b38e2b6f33ef5eaabfdac37e36e8c2dbe5662249257",
"edition_key": "edition-em0026-audit-projection",
"extent": {
"type": "json-value",
"value": {
"automatic_independence_credit": 0,
"dimensions": [
"data",
"method",
"retrieval",
"source_work",
"edition",
"exact_span",
"derivation",
"upstream_citation"
],
"note": "Candidate warrant roots remain connected by declared or unresolved task, retrieval, judge-method, source, edition, span, and derivation lineage."
}
},
"id": "em:dossier-span:sha256:40a70995b7ab3d216ad5a2549fc0ebfacaaf134442733ab984a150b78d487861",
"key": "span-audit-source-boundary",
"locator": {
"label": "Source and derivation dependence boundary",
"pointer": "/source_dependence_boundary",
"type": "json-pointer"
},
"visibility": "public"
}
],
"stage": "draft",
"title": "When eight research agents agree, how many evidence roots are there?",
"visibility": "public"
}
Build receipt
Reproduce this projection
- Catalog
em:catalog:sha256:9bfc972213cba2cde167386103dc2c011ee74639fb7f0794c54120fbbdef1a5d- Frontier
em:frontier:sha256:f33be3eae4c75232d56750ef9a1aa79d96274ece3417d65a75c1391bf61a81bf- Accepted commit
f92846570180dfa4511263f8ba98ecd18f7772c9- Epistemic policy
commons-balanced-v0.1- Disclosure policy
public-noninterference-v0.1- Compiler
epistemedia/0.2.0