{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2026:3Q4F7NJBFDPXY6GCIM4B4UVVO4","short_pith_number":"pith:3Q4F7NJB","schema_version":"1.0","canonical_sha256":"dc385fb52128df7c78c243381e52b5773a1b78a53ac914fe36f34e5b6849543a","source":{"kind":"arxiv","id":"2604.11661","version":3},"attestation_state":"computed","paper":{"title":"Towards Autonomous Mechanistic Reasoning in Virtual Cells","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"Mechanistic action graphs verified by a multi-agent system produce explanations that improve factual accuracy and gene expression prediction in virtual cell models.","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Alisandra Kaye Denton, Dominique Beaini, Emmanuel Noutahi, Jake Fawkes, Lu Zhu, Yunhui Jang","submitted_at":"2026-04-13T16:10:44Z","abstract_excerpt":"Large language models (LLMs) have recently gained significant attention as a promising approach to accelerate scientific discovery. However, their application in open-ended scientific domains such as biology remains limited, primarily due to the lack of factually grounded and actionable explanations. To address this, we introduce a structured explanation formalism for virtual cells that represents biological reasoning as mechanistic action graphs, enabling systematic verification and falsification. Building upon this, we propose VCR-Agent, a multi-agent framework that integrates biologically g"},"verification_status":{"content_addressed":true,"pith_receipt":true,"author_attested":false,"weak_author_claims":0,"strong_author_claims":0,"externally_anchored":false,"storage_verified":false,"citation_signatures":0,"replication_records":0,"graph_snapshot":true,"references_resolved":false,"formal_links_present":false},"canonical_record":{"source":{"id":"2604.11661","kind":"arxiv","version":3},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","primary_cat":"cs.LG","submitted_at":"2026-04-13T16:10:44Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"5c1a0c1236186beafc9239251a4ce6ee0545eec4b92d50194b8326f968f75f83","abstract_canon_sha256":"7c30bdecd5e7233efc149edf3f9b6ef5cd220819eccad51513448d7923768979"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-21T01:05:19.018170Z","signature_b64":"iXHm6qIpoGdeHtTKuGWAU7pOlQXQsEnhbL7TVlx5ZEocdLS5WiaesSeNFRstE779G71gHmHIyAY6oDNDyqEmCw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"dc385fb52128df7c78c243381e52b5773a1b78a53ac914fe36f34e5b6849543a","last_reissued_at":"2026-05-21T01:05:19.017546Z","signature_status":"signed_v1","first_computed_at":"2026-05-21T01:05:19.017546Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Towards Autonomous Mechanistic Reasoning in Virtual Cells","license":"http://creativecommons.org/licenses/by-nc-sa/4.0/","headline":"Mechanistic action graphs verified by a multi-agent system produce explanations that improve factual accuracy and gene expression prediction in virtual cell models.","cross_cats":["cs.AI"],"primary_cat":"cs.LG","authors_text":"Alisandra Kaye Denton, Dominique Beaini, Emmanuel Noutahi, Jake Fawkes, Lu Zhu, Yunhui Jang","submitted_at":"2026-04-13T16:10:44Z","abstract_excerpt":"Large language models (LLMs) have recently gained significant attention as a promising approach to accelerate scientific discovery. However, their application in open-ended scientific domains such as biology remains limited, primarily due to the lack of factually grounded and actionable explanations. To address this, we introduce a structured explanation formalism for virtual cells that represents biological reasoning as mechanistic action graphs, enabling systematic verification and falsification. Building upon this, we propose VCR-Agent, a multi-agent framework that integrates biologically g"},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"training with these explanations improves factual precision and provides a more effective supervision signal for downstream gene expression prediction","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"The verifier-based filtering approach produces factually grounded and actionable mechanistic reasoning that is independent of the generation process and generalizes beyond the Tahoe-100M atlas.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"VCR-Agent uses mechanistic action graphs and verifier-based filtering to autonomously produce verified biological explanations, and training on the resulting VC-TRACES dataset improves factual precision and gene expression prediction.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"Mechanistic action graphs verified by a multi-agent system produce explanations that improve factual accuracy and gene expression prediction in virtual cell models.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"4e6d03dc395d989c2b7f8a21d740c06480c3966d86c1ef069e794d2dd38594de"},"source":{"id":"2604.11661","kind":"arxiv","version":3},"verdict":{"id":"0360602a-4c03-439e-bb1f-91e207111afa","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-10T16:22:58.095350Z","strongest_claim":"training with these explanations improves factual precision and provides a more effective supervision signal for downstream gene expression prediction","one_line_summary":"VCR-Agent uses mechanistic action graphs and verifier-based filtering to autonomously produce verified biological explanations, and training on the resulting VC-TRACES dataset improves factual precision and gene expression prediction.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"The verifier-based filtering approach produces factually grounded and actionable mechanistic reasoning that is independent of the generation process and generalizes beyond the Tahoe-100M atlas.","pith_extraction_headline":"Mechanistic action graphs verified by a multi-agent system produce explanations that improve factual accuracy and gene expression prediction in virtual cell models."},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2604.11661/integrity.json","findings":[],"available":true,"detectors_run":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938"},"references":{"count":0,"sample":[],"resolved_work":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","internal_anchors":0},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"},"aliases":[{"alias_kind":"arxiv","alias_value":"2604.11661","created_at":"2026-05-21T01:05:19.017638+00:00"},{"alias_kind":"arxiv_version","alias_value":"2604.11661v3","created_at":"2026-05-21T01:05:19.017638+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2604.11661","created_at":"2026-05-21T01:05:19.017638+00:00"},{"alias_kind":"pith_short_12","alias_value":"3Q4F7NJBFDPX","created_at":"2026-05-21T01:05:19.017638+00:00"},{"alias_kind":"pith_short_16","alias_value":"3Q4F7NJBFDPXY6GC","created_at":"2026-05-21T01:05:19.017638+00:00"},{"alias_kind":"pith_short_8","alias_value":"3Q4F7NJB","created_at":"2026-05-21T01:05:19.017638+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":0,"internal_anchor_count":0,"sample":[]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/3Q4F7NJBFDPXY6GCIM4B4UVVO4","json":"https://pith.science/pith/3Q4F7NJBFDPXY6GCIM4B4UVVO4.json","graph_json":"https://pith.science/api/pith-number/3Q4F7NJBFDPXY6GCIM4B4UVVO4/graph.json","events_json":"https://pith.science/api/pith-number/3Q4F7NJBFDPXY6GCIM4B4UVVO4/events.json","paper":"https://pith.science/paper/3Q4F7NJB"},"agent_actions":{"view_html":"https://pith.science/pith/3Q4F7NJBFDPXY6GCIM4B4UVVO4","download_json":"https://pith.science/pith/3Q4F7NJBFDPXY6GCIM4B4UVVO4.json","view_paper":"https://pith.science/paper/3Q4F7NJB","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2604.11661&json=true","fetch_graph":"https://pith.science/api/pith-number/3Q4F7NJBFDPXY6GCIM4B4UVVO4/graph.json","fetch_events":"https://pith.science/api/pith-number/3Q4F7NJBFDPXY6GCIM4B4UVVO4/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/3Q4F7NJBFDPXY6GCIM4B4UVVO4/action/timestamp_anchor","attest_storage":"https://pith.science/pith/3Q4F7NJBFDPXY6GCIM4B4UVVO4/action/storage_attestation","attest_author":"https://pith.science/pith/3Q4F7NJBFDPXY6GCIM4B4UVVO4/action/author_attestation","sign_citation":"https://pith.science/pith/3Q4F7NJBFDPXY6GCIM4B4UVVO4/action/citation_signature","submit_replication":"https://pith.science/pith/3Q4F7NJBFDPXY6GCIM4B4UVVO4/action/replication_record"}},"created_at":"2026-05-21T01:05:19.017638+00:00","updated_at":"2026-05-21T01:05:19.017638+00:00"}