{"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"}