{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:USRRMRKMJOZRNWTICZIPBWZ6J2","short_pith_number":"pith:USRRMRKM","schema_version":"1.0","canonical_sha256":"a4a316454c4bb316da681650f0db3e4ea431b5daad76a25260a45ff72209e37a","source":{"kind":"arxiv","id":"2410.17234","version":1},"attestation_state":"computed","paper":{"title":"Fine-Tuning Large Language Models to Appropriately Abstain with Semantic Entropy","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CL","authors_text":"Benedict Aaron Tjandra, Jannik Kossen, Kunal Handa, Muhammed Razzak, Yarin Gal","submitted_at":"2024-10-22T17:54:03Z","abstract_excerpt":"Large Language Models (LLMs) are known to hallucinate, whereby they generate plausible but inaccurate text. This phenomenon poses significant risks in critical applications, such as medicine or law, necessitating robust hallucination mitigation strategies. While recent works have proposed fine-tuning methods to teach LLMs to abstain from answering questions beyond their knowledge or capabilities, these methods rely on the existence of ground-truth labels or are limited to short-form responses. To address these limitations, we propose fine-tuning using semantic entropy, an uncertainty measure d"},"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":"2410.17234","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-10-22T17:54:03Z","cross_cats_sorted":["cs.LG"],"title_canon_sha256":"d5b879eb38a2915b6e0280b011afd6e6fcd92534ff9b37af0cdc6036e2eed3be","abstract_canon_sha256":"92bd6afb9a9956081f96fdb68b54815393a3d5d3fbba056b09978f7893ee2648"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:24:13.922728Z","signature_b64":"SiO3aOxkGPU67jdv4sDzbohJ69xVpMly0AlQ24cLlR3EqVWvO8Lc4Q3Fi9vlwnyL0w4JOr0lvtAYXx6+Z4xDBQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"a4a316454c4bb316da681650f0db3e4ea431b5daad76a25260a45ff72209e37a","last_reissued_at":"2026-07-05T09:24:13.922286Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:24:13.922286Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Fine-Tuning Large Language Models to Appropriately Abstain with Semantic Entropy","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.LG"],"primary_cat":"cs.CL","authors_text":"Benedict Aaron Tjandra, Jannik Kossen, Kunal Handa, Muhammed Razzak, Yarin Gal","submitted_at":"2024-10-22T17:54:03Z","abstract_excerpt":"Large Language Models (LLMs) are known to hallucinate, whereby they generate plausible but inaccurate text. This phenomenon poses significant risks in critical applications, such as medicine or law, necessitating robust hallucination mitigation strategies. While recent works have proposed fine-tuning methods to teach LLMs to abstain from answering questions beyond their knowledge or capabilities, these methods rely on the existence of ground-truth labels or are limited to short-form responses. To address these limitations, we propose fine-tuning using semantic entropy, an uncertainty measure d"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2410.17234","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/2410.17234/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":"2410.17234","created_at":"2026-07-05T09:24:13.922370+00:00"},{"alias_kind":"arxiv_version","alias_value":"2410.17234v1","created_at":"2026-07-05T09:24:13.922370+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2410.17234","created_at":"2026-07-05T09:24:13.922370+00:00"},{"alias_kind":"pith_short_12","alias_value":"USRRMRKMJOZR","created_at":"2026-07-05T09:24:13.922370+00:00"},{"alias_kind":"pith_short_16","alias_value":"USRRMRKMJOZRNWTI","created_at":"2026-07-05T09:24:13.922370+00:00"},{"alias_kind":"pith_short_8","alias_value":"USRRMRKM","created_at":"2026-07-05T09:24:13.922370+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":2,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2603.22161","citing_title":"Causal Evidence that Language Models use Confidence to Drive Behavior","ref_index":19,"is_internal_anchor":false},{"citing_arxiv_id":"2605.12813","citing_title":"REALISTA: Realistic Latent Adversarial Attacks that Elicit LLM Hallucinations","ref_index":84,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/USRRMRKMJOZRNWTICZIPBWZ6J2","json":"https://pith.science/pith/USRRMRKMJOZRNWTICZIPBWZ6J2.json","graph_json":"https://pith.science/api/pith-number/USRRMRKMJOZRNWTICZIPBWZ6J2/graph.json","events_json":"https://pith.science/api/pith-number/USRRMRKMJOZRNWTICZIPBWZ6J2/events.json","paper":"https://pith.science/paper/USRRMRKM"},"agent_actions":{"view_html":"https://pith.science/pith/USRRMRKMJOZRNWTICZIPBWZ6J2","download_json":"https://pith.science/pith/USRRMRKMJOZRNWTICZIPBWZ6J2.json","view_paper":"https://pith.science/paper/USRRMRKM","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2410.17234&json=true","fetch_graph":"https://pith.science/api/pith-number/USRRMRKMJOZRNWTICZIPBWZ6J2/graph.json","fetch_events":"https://pith.science/api/pith-number/USRRMRKMJOZRNWTICZIPBWZ6J2/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/USRRMRKMJOZRNWTICZIPBWZ6J2/action/timestamp_anchor","attest_storage":"https://pith.science/pith/USRRMRKMJOZRNWTICZIPBWZ6J2/action/storage_attestation","attest_author":"https://pith.science/pith/USRRMRKMJOZRNWTICZIPBWZ6J2/action/author_attestation","sign_citation":"https://pith.science/pith/USRRMRKMJOZRNWTICZIPBWZ6J2/action/citation_signature","submit_replication":"https://pith.science/pith/USRRMRKMJOZRNWTICZIPBWZ6J2/action/replication_record"}},"created_at":"2026-07-05T09:24:13.922370+00:00","updated_at":"2026-07-05T09:24:13.922370+00:00"}