{"id":"99b38aada7fcae35a6cfd75096a8e01b0f5a5812bf66c71b6f0211d6e475d0d2","title":"arXiv cs.AI: Tracing the Cascade: A Topology-Aware Evaluation Framework for Scientific Agent Hallucinations","argument":"Authors Xinshun Feng, Ziqi Miao, Lijun Li and Jing Shao introduce SCHEMA, a framework for evaluating hallucinations in scientific agents. It builds scientific concept graphs from benchmark seeds and literature evidence, then creates tasks and checks both intermediate reasoning and selected failures. The authors report that hallucinations cluster around highly connected knowledge hubs, and that correct final answers can result from flawed reasoning. They argue that final-answer accuracy alone is not enough to assess reliability in high-stakes scientific applications. How well does SCHEMA's evaluation work across different scientific fields and agent systems?","topicIds":["intelligence-and-discovery"],"domain":"science","claimType":"attributed reading","limits":"The supplied evidence is the paper's abstract and does not give quantitative results, benchmark scale, or details of how the reported findings vary across scientific fields or agent systems.","speaker":{"name":"Feng, Xinshun, Miao, Ziqi, Li, Lijun, Shao, Jing","roleAtSource":"Article author"},"source":{"title":"Tracing the Cascade: A Topology-Aware Evaluation Framework for Scientific Agent Hallucinations","publisher":"arXiv cs.AI","format":"Article","url":"https://arxiv.org/abs/2608.00711","publishedAt":"2026-08-01","dateBasis":"Date basis not provided."},"evidence":[{"text":"The authors introduce SCHEMA as an evidence-grounded, topology-aware framework for evaluating hallucinations in scientific agents.","excerpt":"We address this gap with SCHEMA, the first evidence-grounded, topology-aware evaluation framework for hallucinations in scientific agents.","url":"https://arxiv.org/abs/2608.00711","relationship":"Source-stated"},{"text":"SCHEMA builds scientific concept graphs from benchmark seeds and literature evidence, and uses them to create several kinds of evaluation tasks.","excerpt":"SCHEMA automatically constructs scientific concept graphs from benchmark seeds and literature evidence, synthesizes graph-grounded tasks spanning claim verification, multi-hop reasoning, open-ended explanation, and experimental code generation","url":"https://arxiv.org/abs/2608.00711","relationship":"Source-stated"},{"text":"The authors report that hallucinations are concentrated around a small number of highly connected knowledge hubs.","excerpt":"SCHEMA reveals that hallucinations concentrate at a small set of highly connected knowledge hubs","url":"https://arxiv.org/abs/2608.00711","relationship":"Source-stated"},{"text":"The authors report that correct final answers can still follow structurally flawed reasoning.","excerpt":"models often reach correct conclusions through structurally flawed reasoning.","url":"https://arxiv.org/abs/2608.00711","relationship":"Source-stated"},{"text":"The authors argue that final-answer accuracy alone is insufficient for judging agent reliability in high-stakes scientific applications.","excerpt":"for high-stakes scientific applications, terminal accuracy alone is an insufficient signal of agent reliability","url":"https://arxiv.org/abs/2608.00711","relationship":"Source-stated"}],"sourceMedia":{"status":"not_found","sourceUrl":"https://arxiv.org/abs/2608.00711","checkedAt":"2026-11-14T00:26:08.292Z","items":[],"note":"No suitable media was found by the supported page extraction."},"assumptions":[],"openQuestion":{"attribution":"Open question","text":"How well does SCHEMA's evaluation work across different scientific fields and agent systems?"},"whatWouldChangeView":{"attribution":"Change condition not recorded","text":""},"editorialComparisons":[],"live":{"title":"arXiv cs.AI: Tracing the Cascade: A Topology-Aware Evaluation Framework for Scientific Agent Hallucinations","summary":"Authors Xinshun Feng, Ziqi Miao, Lijun Li and Jing Shao introduce SCHEMA, a framework for evaluating hallucinations in scientific agents. It builds scientific concept graphs from benchmark seeds and literature evidence, then creates tasks and checks both intermediate reasoning and selected failures. The authors report that hallucinations cluster around highly connected knowledge hubs, and that correct final answers can result from flawed reasoning. They argue that final-answer accuracy alone is not enough to assess reliability in high-stakes scientific applications. How well does SCHEMA's evaluation work across different scientific fields and agent systems?","question":"How well does SCHEMA's evaluation work across different scientific fields and agent systems?","claims":[{"text":"The authors introduce SCHEMA as an evidence-grounded, topology-aware framework for evaluating hallucinations in scientific agents.","excerpt":"We address this gap with SCHEMA, the first evidence-grounded, topology-aware evaluation framework for hallucinations in scientific agents.","support":"source-stated"},{"text":"SCHEMA builds scientific concept graphs from benchmark seeds and literature evidence, and uses them to create several kinds of evaluation tasks.","excerpt":"SCHEMA automatically constructs scientific concept graphs from benchmark seeds and literature evidence, synthesizes graph-grounded tasks spanning claim verification, multi-hop reasoning, open-ended explanation, and experimental code generation","support":"source-stated"},{"text":"The authors report that hallucinations are concentrated around a small number of highly connected knowledge hubs.","excerpt":"SCHEMA reveals that hallucinations concentrate at a small set of highly connected knowledge hubs","support":"source-stated"},{"text":"The authors report that correct final answers can still follow structurally flawed reasoning.","excerpt":"models often reach correct conclusions through structurally flawed reasoning.","support":"source-stated"},{"text":"The authors argue that final-answer accuracy alone is insufficient for judging agent reliability in high-stakes scientific applications.","excerpt":"for high-stakes scientific applications, terminal accuracy alone is an insufficient signal of agent reliability","support":"source-stated"}],"limitations":["The supplied evidence is the paper's abstract and does not give quantitative results, benchmark scale, or details of how the reported findings vary across scientific fields or agent systems."],"mediaId":"328f99be28ff918f343e3cbbaeecee09fa75d46cd221e6a7d638b3f5ad30fb8b","correctionReason":null,"id":"99b38aada7fcae35a6cfd75096a8e01b0f5a5812bf66c71b6f0211d6e475d0d2","publicationId":"5816d45b-3f6f-45be-892e-f6a00ffb1c8d","version":1,"sourceUrl":"https://arxiv.org/abs/2608.00711","sourceTitle":"Tracing the Cascade: A Topology-Aware Evaluation Framework for Scientific Agent Hallucinations","publisher":"arXiv cs.AI","publishedAt":null,"originalPublished":{"value":"2026-08-01","precision":"day","basis":"publisher_citation_date"},"firstObservedAt":"2026-10-28T09:00:39.636Z","createdAt":"2026-11-13T22:18:59.198Z","historical":true,"entityIds":["arxiv"],"topicIds":["intelligence-and-discovery"],"media":[{"id":"328f99be28ff918f343e3cbbaeecee09fa75d46cd221e6a7d638b3f5ad30fb8b","kind":"article","url":"https://arxiv.org/abs/2608.00711","sourceUrl":"https://arxiv.org/abs/2608.00711","label":"Read original","provenance":"source_link","verification":{"status":"verified","checkedAt":"2026-10-30T04:18:19.837Z","method":"acquired_source","contentType":"text/html"},"recommendationReason":"Recommended by Jev for understanding this source; checked by Luna."}],"sourceMedia":{"status":"not_found","sourceUrl":"https://arxiv.org/abs/2608.00711","checkedAt":"2026-11-14T00:26:08.292Z","items":[],"note":"No suitable media was found by the supported page extraction."},"editorialDecision":{"outcome":"accept","reasonCodes":["material_development"],"rubricVersion":"stream-admission-v5","mediaId":"328f99be28ff918f343e3cbbaeecee09fa75d46cd221e6a7d638b3f5ad30fb8b","confidence":0.8,"provider":"typesafe","model":"jev-1.13.0","inputHash":"61f8c3cc851e57bab935145cd1f56b988772c5ccce697854797558466cbc6597","decidedAt":"2026-10-30T04:18:21.507Z"},"provenance":{"textHash":"e0be5b09ce863c9a320d3fd9809dbeb6478c907f1bab7569de6330bbab7319ca","observationId":"70f3c45abe106f684cb27fa9a02a71d8e18b18273cc3db45a52e5de743409628","retrievedAt":"2026-10-30T04:18:19.106Z","kind":"publisher-page","preparer":{"model":"gpt-6-luna","runId":"cf753415-b04e-4ac8-89cb-93b0ca840edc"},"reviewer":{"model":"gpt-6-luna","runId":"4cc4e51c-8651-45e0-855f-d549a390721a"}},"audit":{"status":"uncertain","reviewedAt":"2026-11-13T22:18:59.198Z","model":"anthropic/claude-opus-5.5"},"sourceAttribution":{"authors":[{"name":"Feng, Xinshun","url":null,"kind":"unknown"},{"name":"Miao, Ziqi","url":null,"kind":"unknown"},{"name":"Li, Lijun","url":null,"kind":"unknown"},{"name":"Shao, Jing","url":null,"kind":"unknown"}],"authorRole":"article-author","hosts":[],"status":"stated","authorRoster":{"truncated":false,"omitted":0,"assertionGroupsOmitted":0},"participation":[{"actor":{"name":"Feng, Xinshun","url":null,"kind":"unknown","actorId":null},"role":"author","scope":{"kind":"work-credit","locator":null},"literalCredit":"Feng, Xinshun","status":"stated","provenance":{"basis":"head-meta","sourceUrl":"https://arxiv.org/abs/2608.00711","locator":"head/meta[name=author|citation_author]"},"assertionId":"15a7ef75ca2f2507cdb9a3703c33efacc7ef35c020e1f6461a06e9b5076ef49f:participation:0","identity":{"version":"entity-identity-v1","status":"unresolved","actorId":null,"canonicalActorId":null,"candidateActorIds":[],"assertions":[],"sourceObservedAt":"2026-11-14T00:26:08.292Z","knowledgeAsOf":null},"sourceVersion":{"snapshotId":"15a7ef75ca2f2507cdb9a3703c33efacc7ef35c020e1f6461a06e9b5076ef49f","sourceUrl":"https://arxiv.org/abs/2608.00711","contentSha256":"0fe09c8b7b46634810408ed9523058a4712777d5ffe95a9c9dbecf9c1d75acdb","textHash":"e0be5b09ce863c9a320d3fd9809dbeb6478c907f1bab7569de6330bbab7319ca","observedAt":"2026-11-14T00:26:08.292Z"}},{"actor":{"name":"Miao, Ziqi","url":null,"kind":"unknown","actorId":null},"role":"author","scope":{"kind":"work-credit","locator":null},"literalCredit":"Miao, Ziqi","status":"stated","provenance":{"basis":"head-meta","sourceUrl":"https://arxiv.org/abs/2608.00711","locator":"head/meta[name=author|citation_author]"},"assertionId":"15a7ef75ca2f2507cdb9a3703c33efacc7ef35c020e1f6461a06e9b5076ef49f:participation:1","identity":{"version":"entity-identity-v1","status":"unresolved","actorId":null,"canonicalActorId":null,"candidateActorIds":[],"assertions":[],"sourceObservedAt":"2026-11-14T00:26:08.292Z","knowledgeAsOf":null},"sourceVersion":{"snapshotId":"15a7ef75ca2f2507cdb9a3703c33efacc7ef35c020e1f6461a06e9b5076ef49f","sourceUrl":"https://arxiv.org/abs/2608.00711","contentSha256":"0fe09c8b7b46634810408ed9523058a4712777d5ffe95a9c9dbecf9c1d75acdb","textHash":"e0be5b09ce863c9a320d3fd9809dbeb6478c907f1bab7569de6330bbab7319ca","observedAt":"2026-11-14T00:26:08.292Z"}},{"actor":{"name":"Li, Lijun","url":null,"kind":"unknown","actorId":null},"role":"author","scope":{"kind":"work-credit","locator":null},"literalCredit":"Li, Lijun","status":"stated","provenance":{"basis":"head-meta","sourceUrl":"https://arxiv.org/abs/2608.00711","locator":"head/meta[name=author|citation_author]"},"assertionId":"15a7ef75ca2f2507cdb9a3703c33efacc7ef35c020e1f6461a06e9b5076ef49f:participation:2","identity":{"version":"entity-identity-v1","status":"unresolved","actorId":null,"canonicalActorId":null,"candidateActorIds":[],"assertions":[],"sourceObservedAt":"2026-11-14T00:26:08.292Z","knowledgeAsOf":null},"sourceVersion":{"snapshotId":"15a7ef75ca2f2507cdb9a3703c33efacc7ef35c020e1f6461a06e9b5076ef49f","sourceUrl":"https://arxiv.org/abs/2608.00711","contentSha256":"0fe09c8b7b46634810408ed9523058a4712777d5ffe95a9c9dbecf9c1d75acdb","textHash":"e0be5b09ce863c9a320d3fd9809dbeb6478c907f1bab7569de6330bbab7319ca","observedAt":"2026-11-14T00:26:08.292Z"}},{"actor":{"name":"Shao, Jing","url":null,"kind":"unknown","actorId":null},"role":"author","scope":{"kind":"work-credit","locator":null},"literalCredit":"Shao, Jing","status":"stated","provenance":{"basis":"head-meta","sourceUrl":"https://arxiv.org/abs/2608.00711","locator":"head/meta[name=author|citation_author]"},"assertionId":"15a7ef75ca2f2507cdb9a3703c33efacc7ef35c020e1f6461a06e9b5076ef49f:participation:3","identity":{"version":"entity-identity-v1","status":"unresolved","actorId":null,"canonicalActorId":null,"candidateActorIds":[],"assertions":[],"sourceObservedAt":"2026-11-14T00:26:08.292Z","knowledgeAsOf":null},"sourceVersion":{"snapshotId":"15a7ef75ca2f2507cdb9a3703c33efacc7ef35c020e1f6461a06e9b5076ef49f","sourceUrl":"https://arxiv.org/abs/2608.00711","contentSha256":"0fe09c8b7b46634810408ed9523058a4712777d5ffe95a9c9dbecf9c1d75acdb","textHash":"e0be5b09ce863c9a320d3fd9809dbeb6478c907f1bab7569de6330bbab7319ca","observedAt":"2026-11-14T00:26:08.292Z"}}],"participationCoverage":{"total":4,"returned":4,"complete":true,"fullAssertions":"source-metadata-endpoint"},"distribution":null,"publisher":null,"sourceUrl":"https://arxiv.org/abs/2608.00711","snapshotId":"15a7ef75ca2f2507cdb9a3703c33efacc7ef35c020e1f6461a06e9b5076ef49f","contentSha256":"0fe09c8b7b46634810408ed9523058a4712777d5ffe95a9c9dbecf9c1d75acdb","textHash":"e0be5b09ce863c9a320d3fd9809dbeb6478c907f1bab7569de6330bbab7319ca","observedAt":"2026-11-14T00:26:08.292Z","basis":"publisher-page-metadata","sourceLinks":45,"linksTruncated":false,"metadataUrl":"/api/pipeline/stream/99b38aada7fcae35a6cfd75096a8e01b0f5a5812bf66c71b6f0211d6e475d0d2/source"}},"canonicalUrl":"https://escapemeridian.com/readings/arxiv-cs-ai-tracing-the-cascade-a-topology-aware-evaluation-99b38aad","formats":{"html":"/readings/arxiv-cs-ai-tracing-the-cascade-a-topology-aware-evaluation-99b38aad","markdown":"/readings/arxiv-cs-ai-tracing-the-cascade-a-topology-aware-evaluation-99b38aad.md","json":"/readings/arxiv-cs-ai-tracing-the-cascade-a-topology-aware-evaluation-99b38aad.json"},"attribution":"Headline and summary by Escape Meridian, prepared by Luna from the attributed source. A source check is not independent confirmation of the source’s claims.","maintainedGuideLinks":[],"conditions":[],"readerContext":{"version":"reader-context-v1","attribution":"Reference context selected by Escape Meridian; separate from source evidence. Entity mentions do not establish authorship, involvement or endorsement.","terms":[{"id":"term-hallucination","name":"Hallucination","explanation":"When an AI system states something false or made up as if it were true.","referenceUrl":"https://developers.google.com/machine-learning/glossary#hallucination"},{"id":"term-benchmark","name":"Benchmark","explanation":"A defined set of tests used to compare systems. Results depend on the tasks and measures selected.","referenceUrl":"https://arxiv.org/abs/2211.09110"}],"entities":[{"id":"arxiv","name":"arXiv","kind":"organization","context":"arXiv is a research manuscript repository where authors share papers that may not yet have been peer reviewed.","referenceUrls":["https://arxiv.org/","https://info.arxiv.org/help/api/index.html"],"path":"/entities/arxiv","json":"/entities/arxiv.json","markdown":"/entities/arxiv.md"}]},"researchQuestions":[],"withdrawn":false,"references":{"schemaVersion":1,"target":{"type":"reading","id":"99b38aada7fcae35a6cfd75096a8e01b0f5a5812bf66c71b6f0211d6e475d0d2","revision":"5816d45b-3f6f-45be-892e-f6a00ffb1c8d","revisionKind":"publication-id","title":"arXiv cs.AI: Tracing the Cascade: A Topology-Aware Evaluation Framework for Scientific Agent Hallucinations","url":"https://escapemeridian.com/readings/arxiv-cs-ai-tracing-the-cascade-a-topology-aware-evaluation-99b38aad"},"source":{"type":"source","id":"https://arxiv.org/abs/2608.00711","revision":"e0be5b09ce863c9a320d3fd9809dbeb6478c907f1bab7569de6330bbab7319ca","revisionKind":"sha256-source-text","title":"Tracing the Cascade: A Topology-Aware Evaluation Framework for Scientific Agent Hallucinations","url":"https://arxiv.org/abs/2608.00711"},"status":"available","items":[],"next":null,"scope":"Explicit citations in public readings and public file attachments to this reading. Shared topics are not citations. Other artifact types need an explicit reference adapter."},"revisions":[{"id":"99b38aada7fcae35a6cfd75096a8e01b0f5a5812bf66c71b6f0211d6e475d0d2","title":"arXiv cs.AI: Tracing the Cascade: A Topology-Aware Evaluation Framework for Scientific Agent Hallucinations","argument":"Authors Xinshun Feng, Ziqi Miao, Lijun Li and Jing Shao introduce SCHEMA, a framework for evaluating hallucinations in scientific agents. It builds scientific concept graphs from benchmark seeds and literature evidence, then creates tasks and checks both intermediate reasoning and selected failures. The authors report that hallucinations cluster around highly connected knowledge hubs, and that correct final answers can result from flawed reasoning. They argue that final-answer accuracy alone is not enough to assess reliability in high-stakes scientific applications. How well does SCHEMA's evaluation work across different scientific fields and agent systems?","topicIds":["intelligence-and-discovery"],"domain":"science","claimType":"attributed reading","limits":"The supplied evidence is the paper's abstract and does not give quantitative results, benchmark scale, or details of how the reported findings vary across scientific fields or agent systems.","speaker":{"name":"Feng, Xinshun, Miao, Ziqi, Li, Lijun, Shao, Jing","roleAtSource":"Article author"},"source":{"title":"Tracing the Cascade: A Topology-Aware Evaluation Framework for Scientific Agent Hallucinations","publisher":"arXiv cs.AI","format":"Article","url":"https://arxiv.org/abs/2608.00711","publishedAt":"2026-08-01","dateBasis":"Date basis not provided."},"evidence":[{"text":"The authors introduce SCHEMA as an evidence-grounded, topology-aware framework for evaluating hallucinations in scientific agents.","excerpt":"We address this gap with SCHEMA, the first evidence-grounded, topology-aware evaluation framework for hallucinations in scientific agents.","url":"https://arxiv.org/abs/2608.00711","relationship":"Source-stated"},{"text":"SCHEMA builds scientific concept graphs from benchmark seeds and literature evidence, and uses them to create several kinds of evaluation tasks.","excerpt":"SCHEMA automatically constructs scientific concept graphs from benchmark seeds and literature evidence, synthesizes graph-grounded tasks spanning claim verification, multi-hop reasoning, open-ended explanation, and experimental code generation","url":"https://arxiv.org/abs/2608.00711","relationship":"Source-stated"},{"text":"The authors report that hallucinations are concentrated around a small number of highly connected knowledge hubs.","excerpt":"SCHEMA reveals that hallucinations concentrate at a small set of highly connected knowledge hubs","url":"https://arxiv.org/abs/2608.00711","relationship":"Source-stated"},{"text":"The authors report that correct final answers can still follow structurally flawed reasoning.","excerpt":"models often reach correct conclusions through structurally flawed reasoning.","url":"https://arxiv.org/abs/2608.00711","relationship":"Source-stated"},{"text":"The authors argue that final-answer accuracy alone is insufficient for judging agent reliability in high-stakes scientific applications.","excerpt":"for high-stakes scientific applications, terminal accuracy alone is an insufficient signal of agent reliability","url":"https://arxiv.org/abs/2608.00711","relationship":"Source-stated"}],"sourceMedia":{"status":"not_found","sourceUrl":"https://arxiv.org/abs/2608.00711","checkedAt":"2026-11-14T00:26:08.292Z","items":[],"note":"No suitable media was found by the supported page extraction."},"assumptions":[],"openQuestion":{"attribution":"Open question","text":"How well does SCHEMA's evaluation work across different scientific fields and agent systems?"},"whatWouldChangeView":{"attribution":"Change condition not recorded","text":""},"editorialComparisons":[],"live":{"title":"arXiv cs.AI: Tracing the Cascade: A Topology-Aware Evaluation Framework for Scientific Agent Hallucinations","summary":"Authors Xinshun Feng, Ziqi Miao, Lijun Li and Jing Shao introduce SCHEMA, a framework for evaluating hallucinations in scientific agents. It builds scientific concept graphs from benchmark seeds and literature evidence, then creates tasks and checks both intermediate reasoning and selected failures. The authors report that hallucinations cluster around highly connected knowledge hubs, and that correct final answers can result from flawed reasoning. They argue that final-answer accuracy alone is not enough to assess reliability in high-stakes scientific applications. How well does SCHEMA's evaluation work across different scientific fields and agent systems?","question":"How well does SCHEMA's evaluation work across different scientific fields and agent systems?","claims":[{"text":"The authors introduce SCHEMA as an evidence-grounded, topology-aware framework for evaluating hallucinations in scientific agents.","excerpt":"We address this gap with SCHEMA, the first evidence-grounded, topology-aware evaluation framework for hallucinations in scientific agents.","support":"source-stated"},{"text":"SCHEMA builds scientific concept graphs from benchmark seeds and literature evidence, and uses them to create several kinds of evaluation tasks.","excerpt":"SCHEMA automatically constructs scientific concept graphs from benchmark seeds and literature evidence, synthesizes graph-grounded tasks spanning claim verification, multi-hop reasoning, open-ended explanation, and experimental code generation","support":"source-stated"},{"text":"The authors report that hallucinations are concentrated around a small number of highly connected knowledge hubs.","excerpt":"SCHEMA reveals that hallucinations concentrate at a small set of highly connected knowledge hubs","support":"source-stated"},{"text":"The authors report that correct final answers can still follow structurally flawed reasoning.","excerpt":"models often reach correct conclusions through structurally flawed reasoning.","support":"source-stated"},{"text":"The authors argue that final-answer accuracy alone is insufficient for judging agent reliability in high-stakes scientific applications.","excerpt":"for high-stakes scientific applications, terminal accuracy alone is an insufficient signal of agent reliability","support":"source-stated"}],"limitations":["The supplied evidence is the paper's abstract and does not give quantitative results, benchmark scale, or details of how the reported findings vary across scientific fields or agent systems."],"mediaId":"328f99be28ff918f343e3cbbaeecee09fa75d46cd221e6a7d638b3f5ad30fb8b","correctionReason":null,"id":"99b38aada7fcae35a6cfd75096a8e01b0f5a5812bf66c71b6f0211d6e475d0d2","publicationId":"5816d45b-3f6f-45be-892e-f6a00ffb1c8d","version":1,"sourceUrl":"https://arxiv.org/abs/2608.00711","sourceTitle":"Tracing the Cascade: A Topology-Aware Evaluation Framework for Scientific Agent Hallucinations","publisher":"arXiv cs.AI","publishedAt":null,"originalPublished":{"value":"2026-08-01","precision":"day","basis":"publisher_citation_date"},"firstObservedAt":"2026-10-28T09:00:39.636Z","createdAt":"2026-11-13T22:18:59.198Z","historical":true,"entityIds":["arxiv"],"topicIds":["intelligence-and-discovery"],"media":[{"id":"328f99be28ff918f343e3cbbaeecee09fa75d46cd221e6a7d638b3f5ad30fb8b","kind":"article","url":"https://arxiv.org/abs/2608.00711","sourceUrl":"https://arxiv.org/abs/2608.00711","label":"Read original","provenance":"source_link","verification":{"status":"verified","checkedAt":"2026-10-30T04:18:19.837Z","method":"acquired_source","contentType":"text/html"},"recommendationReason":"Recommended by Jev for understanding this source; checked by Luna."}],"sourceMedia":{"status":"not_found","sourceUrl":"https://arxiv.org/abs/2608.00711","checkedAt":"2026-11-14T00:26:08.292Z","items":[],"note":"No suitable media was found by the supported page extraction."},"editorialDecision":{"outcome":"accept","reasonCodes":["material_development"],"rubricVersion":"stream-admission-v5","mediaId":"328f99be28ff918f343e3cbbaeecee09fa75d46cd221e6a7d638b3f5ad30fb8b","confidence":0.8,"provider":"typesafe","model":"jev-1.13.0","inputHash":"61f8c3cc851e57bab935145cd1f56b988772c5ccce697854797558466cbc6597","decidedAt":"2026-10-30T04:18:21.507Z"},"provenance":{"textHash":"e0be5b09ce863c9a320d3fd9809dbeb6478c907f1bab7569de6330bbab7319ca","observationId":"70f3c45abe106f684cb27fa9a02a71d8e18b18273cc3db45a52e5de743409628","retrievedAt":"2026-10-30T04:18:19.106Z","kind":"publisher-page","preparer":{"model":"gpt-6-luna","runId":"cf753415-b04e-4ac8-89cb-93b0ca840edc"},"reviewer":{"model":"gpt-6-luna","runId":"4cc4e51c-8651-45e0-855f-d549a390721a"}},"audit":{"status":"uncertain","reviewedAt":"2026-11-13T22:18:59.198Z","model":"anthropic/claude-opus-5.5"},"sourceAttribution":{"authors":[{"name":"Feng, Xinshun","url":null,"kind":"unknown"},{"name":"Miao, Ziqi","url":null,"kind":"unknown"},{"name":"Li, Lijun","url":null,"kind":"unknown"},{"name":"Shao, Jing","url":null,"kind":"unknown"}],"authorRole":"article-author","hosts":[],"status":"stated","authorRoster":{"truncated":false,"omitted":0,"assertionGroupsOmitted":0},"participation":[{"actor":{"name":"Feng, Xinshun","url":null,"kind":"unknown","actorId":null},"role":"author","scope":{"kind":"work-credit","locator":null},"literalCredit":"Feng, Xinshun","status":"stated","provenance":{"basis":"head-meta","sourceUrl":"https://arxiv.org/abs/2608.00711","locator":"head/meta[name=author|citation_author]"},"assertionId":"15a7ef75ca2f2507cdb9a3703c33efacc7ef35c020e1f6461a06e9b5076ef49f:participation:0","identity":{"version":"entity-identity-v1","status":"unresolved","actorId":null,"canonicalActorId":null,"candidateActorIds":[],"assertions":[],"sourceObservedAt":"2026-11-14T00:26:08.292Z","knowledgeAsOf":null},"sourceVersion":{"snapshotId":"15a7ef75ca2f2507cdb9a3703c33efacc7ef35c020e1f6461a06e9b5076ef49f","sourceUrl":"https://arxiv.org/abs/2608.00711","contentSha256":"0fe09c8b7b46634810408ed9523058a4712777d5ffe95a9c9dbecf9c1d75acdb","textHash":"e0be5b09ce863c9a320d3fd9809dbeb6478c907f1bab7569de6330bbab7319ca","observedAt":"2026-11-14T00:26:08.292Z"}},{"actor":{"name":"Miao, Ziqi","url":null,"kind":"unknown","actorId":null},"role":"author","scope":{"kind":"work-credit","locator":null},"literalCredit":"Miao, Ziqi","status":"stated","provenance":{"basis":"head-meta","sourceUrl":"https://arxiv.org/abs/2608.00711","locator":"head/meta[name=author|citation_author]"},"assertionId":"15a7ef75ca2f2507cdb9a3703c33efacc7ef35c020e1f6461a06e9b5076ef49f:participation:1","identity":{"version":"entity-identity-v1","status":"unresolved","actorId":null,"canonicalActorId":null,"candidateActorIds":[],"assertions":[],"sourceObservedAt":"2026-11-14T00:26:08.292Z","knowledgeAsOf":null},"sourceVersion":{"snapshotId":"15a7ef75ca2f2507cdb9a3703c33efacc7ef35c020e1f6461a06e9b5076ef49f","sourceUrl":"https://arxiv.org/abs/2608.00711","contentSha256":"0fe09c8b7b46634810408ed9523058a4712777d5ffe95a9c9dbecf9c1d75acdb","textHash":"e0be5b09ce863c9a320d3fd9809dbeb6478c907f1bab7569de6330bbab7319ca","observedAt":"2026-11-14T00:26:08.292Z"}},{"actor":{"name":"Li, Lijun","url":null,"kind":"unknown","actorId":null},"role":"author","scope":{"kind":"work-credit","locator":null},"literalCredit":"Li, Lijun","status":"stated","provenance":{"basis":"head-meta","sourceUrl":"https://arxiv.org/abs/2608.00711","locator":"head/meta[name=author|citation_author]"},"assertionId":"15a7ef75ca2f2507cdb9a3703c33efacc7ef35c020e1f6461a06e9b5076ef49f:participation:2","identity":{"version":"entity-identity-v1","status":"unresolved","actorId":null,"canonicalActorId":null,"candidateActorIds":[],"assertions":[],"sourceObservedAt":"2026-11-14T00:26:08.292Z","knowledgeAsOf":null},"sourceVersion":{"snapshotId":"15a7ef75ca2f2507cdb9a3703c33efacc7ef35c020e1f6461a06e9b5076ef49f","sourceUrl":"https://arxiv.org/abs/2608.00711","contentSha256":"0fe09c8b7b46634810408ed9523058a4712777d5ffe95a9c9dbecf9c1d75acdb","textHash":"e0be5b09ce863c9a320d3fd9809dbeb6478c907f1bab7569de6330bbab7319ca","observedAt":"2026-11-14T00:26:08.292Z"}},{"actor":{"name":"Shao, Jing","url":null,"kind":"unknown","actorId":null},"role":"author","scope":{"kind":"work-credit","locator":null},"literalCredit":"Shao, Jing","status":"stated","provenance":{"basis":"head-meta","sourceUrl":"https://arxiv.org/abs/2608.00711","locator":"head/meta[name=author|citation_author]"},"assertionId":"15a7ef75ca2f2507cdb9a3703c33efacc7ef35c020e1f6461a06e9b5076ef49f:participation:3","identity":{"version":"entity-identity-v1","status":"unresolved","actorId":null,"canonicalActorId":null,"candidateActorIds":[],"assertions":[],"sourceObservedAt":"2026-11-14T00:26:08.292Z","knowledgeAsOf":null},"sourceVersion":{"snapshotId":"15a7ef75ca2f2507cdb9a3703c33efacc7ef35c020e1f6461a06e9b5076ef49f","sourceUrl":"https://arxiv.org/abs/2608.00711","contentSha256":"0fe09c8b7b46634810408ed9523058a4712777d5ffe95a9c9dbecf9c1d75acdb","textHash":"e0be5b09ce863c9a320d3fd9809dbeb6478c907f1bab7569de6330bbab7319ca","observedAt":"2026-11-14T00:26:08.292Z"}}],"participationCoverage":{"total":4,"returned":4,"complete":true,"fullAssertions":"source-metadata-endpoint"},"distribution":null,"publisher":null,"sourceUrl":"https://arxiv.org/abs/2608.00711","snapshotId":"15a7ef75ca2f2507cdb9a3703c33efacc7ef35c020e1f6461a06e9b5076ef49f","contentSha256":"0fe09c8b7b46634810408ed9523058a4712777d5ffe95a9c9dbecf9c1d75acdb","textHash":"e0be5b09ce863c9a320d3fd9809dbeb6478c907f1bab7569de6330bbab7319ca","observedAt":"2026-11-14T00:26:08.292Z","basis":"publisher-page-metadata","sourceLinks":45,"linksTruncated":false,"metadataUrl":"/api/pipeline/stream/99b38aada7fcae35a6cfd75096a8e01b0f5a5812bf66c71b6f0211d6e475d0d2/source"}}}],"audits":[{"publicationId":"5816d45b-3f6f-45be-892e-f6a00ffb1c8d","verdict":"uncertain","findings":[{"dimension":"attribution","status":"pass","explanation":"The reading correctly credits Xinshun Feng, Ziqi Miao, Lijun Li and Jing Shao for SCHEMA. Findings and conclusions stay framed as what the authors report or argue, not as established fact.","excerpt":"Authors:Xinshun Feng, Ziqi Miao, Lijun Li, Jing Shao"},{"dimension":"summary","status":"uncertain","explanation":"The summary omits the paper's central motivation, the cascade named in the title: one wrong claim about a foundational concept can spread through multi-step reasoning and corrupt the whole chain. Without that point, a non-specialist cannot see why clustering at hubs matters. It also softens 'often' to 'can'. 'Hallucinations', 'scientific agents' and 'knowledge hubs' go unexplained, and the title only repeats the paper title. Smallest repair: add one sentence on the propagation mechanism with a plain gloss of an agent and a hallucination, and restore 'often'.","excerpt":"a single erroneous claim on a foundational concept can propagate through multi-step reasoning and corrupt entire trajectories"},{"dimension":"date","status":"pass","explanation":"The originalPublished date of 2026-08-01 matches the v1 submission date. It does not use the 27 Oct 2026 v2 revision or the October feed date.","excerpt":"Submitted on 1 Aug 2026 (v1), last revised 27 Oct 2026 (this version, v2)"},{"dimension":"support","status":"pass","explanation":"Each claim is directly supported by an abstract excerpt and labeled source-stated. The limitation correctly notes that the evidence gives no quantitative results. The question about generalization across fields and agent systems follows from that gap without presuming an answer. No benefits or costs are invented.","excerpt":"These results indicate that for high-stakes scientific applications, terminal accuracy alone is an insufficient signal of agent reliability"},{"dimension":"media","status":"pass","explanation":"The selected media is the verified direct arXiv abstract page for this paper. It is the original source and the most direct access point.","excerpt":null}],"createdAt":"2026-11-13T22:18:59.198Z","model":"anthropic/claude-opus-5.5"}],"correctionUrl":"https://escapemeridian.com/corrections/"}