{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2025:CT6T5OCBJUJZLMIP47VUBC6BLL","short_pith_number":"pith:CT6T5OCB","schema_version":"1.0","canonical_sha256":"14fd3eb8414d1395b10fe7eb408bc15acf2acaa525ce7939579f0941d891862b","source":{"kind":"arxiv","id":"2509.02401","version":1},"attestation_state":"computed","paper":{"title":"Towards Agents That Know When They Don't Know: Uncertainty as a Control Signal for Structured Reasoning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Gianluca Mazzoni, Jesper Ferkinghoff-Borg, Josefa Lia Stoisser, Julien Fauqueur, Kaspar Martens, Lawrence Phillips, Lea M{\\o}rch Harder, Marc Boubnovski Martell, Philip Torr","submitted_at":"2025-09-02T15:12:10Z","abstract_excerpt":"Large language model (LLM) agents are increasingly deployed in structured biomedical data environments, yet they often produce fluent but overconfident outputs when reasoning over complex multi-table data. We introduce an uncertainty-aware agent for query-conditioned multi-table summarization that leverages two complementary signals: (i) retrieval uncertainty--entropy over multiple table-selection rollouts--and (ii) summary uncertainty--combining self-consistency and perplexity. Summary uncertainty is incorporated into reinforcement learning (RL) with Group Relative Policy Optimization (GRPO),"},"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":"2509.02401","kind":"arxiv","version":1},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.AI","submitted_at":"2025-09-02T15:12:10Z","cross_cats_sorted":[],"title_canon_sha256":"962514e00aab31c4975bf1cbe2c21b13a1e4a75039b92fa9d0bb9ab30aedbd1f","abstract_canon_sha256":"77687adf252a547c3eea3f6e996e9df0f844fdf55ed5d5e1ad21fcca6bfa024d"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T12:03:37.910296Z","signature_b64":"xJElSc/H0wsWwCoPkdmOKVNHeURC30VJUzWAc48iHQjpszirvFTyfm7atVpYrIWOk7cPyNEN2wmkkIuukIXRDg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"14fd3eb8414d1395b10fe7eb408bc15acf2acaa525ce7939579f0941d891862b","last_reissued_at":"2026-07-05T12:03:37.909720Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T12:03:37.909720Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Towards Agents That Know When They Don't Know: Uncertainty as a Control Signal for Structured Reasoning","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.AI","authors_text":"Gianluca Mazzoni, Jesper Ferkinghoff-Borg, Josefa Lia Stoisser, Julien Fauqueur, Kaspar Martens, Lawrence Phillips, Lea M{\\o}rch Harder, Marc Boubnovski Martell, Philip Torr","submitted_at":"2025-09-02T15:12:10Z","abstract_excerpt":"Large language model (LLM) agents are increasingly deployed in structured biomedical data environments, yet they often produce fluent but overconfident outputs when reasoning over complex multi-table data. We introduce an uncertainty-aware agent for query-conditioned multi-table summarization that leverages two complementary signals: (i) retrieval uncertainty--entropy over multiple table-selection rollouts--and (ii) summary uncertainty--combining self-consistency and perplexity. Summary uncertainty is incorporated into reinforcement learning (RL) with Group Relative Policy Optimization (GRPO),"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2509.02401","kind":"arxiv","version":1},"verdict":{"id":null,"model_set":{},"created_at":null,"strongest_claim":"","one_line_summary":"","pipeline_version":null,"weakest_assumption":"","pith_extraction_headline":""},"integrity":{"clean":true,"summary":{"advisory":0,"critical":0,"by_detector":{},"informational":0},"endpoint":"/pith/2509.02401/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":"2509.02401","created_at":"2026-07-05T12:03:37.909780+00:00"},{"alias_kind":"arxiv_version","alias_value":"2509.02401v1","created_at":"2026-07-05T12:03:37.909780+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2509.02401","created_at":"2026-07-05T12:03:37.909780+00:00"},{"alias_kind":"pith_short_12","alias_value":"CT6T5OCBJUJZ","created_at":"2026-07-05T12:03:37.909780+00:00"},{"alias_kind":"pith_short_16","alias_value":"CT6T5OCBJUJZLMIP","created_at":"2026-07-05T12:03:37.909780+00:00"},{"alias_kind":"pith_short_8","alias_value":"CT6T5OCB","created_at":"2026-07-05T12:03:37.909780+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.20662","citing_title":"Confidence Laundering in Agent Systems: Why Uncertainty Needs a Latent Carrier","ref_index":41,"is_internal_anchor":false},{"citing_arxiv_id":"2604.12632","citing_title":"Calibration-Aware Policy Optimization for Reasoning LLMs","ref_index":29,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/CT6T5OCBJUJZLMIP47VUBC6BLL","json":"https://pith.science/pith/CT6T5OCBJUJZLMIP47VUBC6BLL.json","graph_json":"https://pith.science/api/pith-number/CT6T5OCBJUJZLMIP47VUBC6BLL/graph.json","events_json":"https://pith.science/api/pith-number/CT6T5OCBJUJZLMIP47VUBC6BLL/events.json","paper":"https://pith.science/paper/CT6T5OCB"},"agent_actions":{"view_html":"https://pith.science/pith/CT6T5OCBJUJZLMIP47VUBC6BLL","download_json":"https://pith.science/pith/CT6T5OCBJUJZLMIP47VUBC6BLL.json","view_paper":"https://pith.science/paper/CT6T5OCB","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2509.02401&json=true","fetch_graph":"https://pith.science/api/pith-number/CT6T5OCBJUJZLMIP47VUBC6BLL/graph.json","fetch_events":"https://pith.science/api/pith-number/CT6T5OCBJUJZLMIP47VUBC6BLL/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/CT6T5OCBJUJZLMIP47VUBC6BLL/action/timestamp_anchor","attest_storage":"https://pith.science/pith/CT6T5OCBJUJZLMIP47VUBC6BLL/action/storage_attestation","attest_author":"https://pith.science/pith/CT6T5OCBJUJZLMIP47VUBC6BLL/action/author_attestation","sign_citation":"https://pith.science/pith/CT6T5OCBJUJZLMIP47VUBC6BLL/action/citation_signature","submit_replication":"https://pith.science/pith/CT6T5OCBJUJZLMIP47VUBC6BLL/action/replication_record"}},"created_at":"2026-07-05T12:03:37.909780+00:00","updated_at":"2026-07-05T12:03:37.909780+00:00"}