{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:J3CY3Q7K2Z3N7ROO7J7UNCLU3A","short_pith_number":"pith:J3CY3Q7K","schema_version":"1.0","canonical_sha256":"4ec58dc3ead676dfc5cefa7f468974d80d8e4e97689e81889732e70116f157f0","source":{"kind":"arxiv","id":"2402.06544","version":2},"attestation_state":"computed","paper":{"title":"Calibrating Long-form Generations from Large Language Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.CL","authors_text":"Arman Cohan, Bhuwan Dhingra, Raghuveer Thirukovalluru, Yixin Liu, Yukun Huang","submitted_at":"2024-02-09T17:00:32Z","abstract_excerpt":"To enhance Large Language Models' (LLMs) reliability, calibration is essential -- the model's assessed confidence scores should align with the actual likelihood of its responses being correct. However, current confidence elicitation methods and calibration metrics typically rely on a binary true/false assessment of response correctness. This approach does not apply to long-form generation, where an answer can be partially correct. Addressing this gap, we introduce a unified calibration framework, in which both the correctness of the LLMs' responses and their associated confidence levels are tr"},"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":"2402.06544","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-02-09T17:00:32Z","cross_cats_sorted":["cs.AI","cs.LG"],"title_canon_sha256":"d81e7e187964d1bda05ca9a9225c110ff3d2ffdf3a05e0634ed5d55d7ca88a8d","abstract_canon_sha256":"9fb670346ead3789426f466f72f90347ac5c9e0a7f973f6b83463647e48b65b5"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T09:26:24.254824Z","signature_b64":"Jkgr+/iPmnNgTuZJ73rppr4a6KW4E5aNpowiUp3UsyQK2RQN1zGUvcxVzpHlPwUgx9yRR1FBR9UVmFndtXgcAg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"4ec58dc3ead676dfc5cefa7f468974d80d8e4e97689e81889732e70116f157f0","last_reissued_at":"2026-07-05T09:26:24.253627Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T09:26:24.253627Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Calibrating Long-form Generations from Large Language Models","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.CL","authors_text":"Arman Cohan, Bhuwan Dhingra, Raghuveer Thirukovalluru, Yixin Liu, Yukun Huang","submitted_at":"2024-02-09T17:00:32Z","abstract_excerpt":"To enhance Large Language Models' (LLMs) reliability, calibration is essential -- the model's assessed confidence scores should align with the actual likelihood of its responses being correct. However, current confidence elicitation methods and calibration metrics typically rely on a binary true/false assessment of response correctness. This approach does not apply to long-form generation, where an answer can be partially correct. Addressing this gap, we introduce a unified calibration framework, in which both the correctness of the LLMs' responses and their associated confidence levels are tr"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2402.06544","kind":"arxiv","version":2},"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/2402.06544/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":"2402.06544","created_at":"2026-07-05T09:26:24.254338+00:00"},{"alias_kind":"arxiv_version","alias_value":"2402.06544v2","created_at":"2026-07-05T09:26:24.254338+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2402.06544","created_at":"2026-07-05T09:26:24.254338+00:00"},{"alias_kind":"pith_short_12","alias_value":"J3CY3Q7K2Z3N","created_at":"2026-07-05T09:26:24.254338+00:00"},{"alias_kind":"pith_short_16","alias_value":"J3CY3Q7K2Z3N7ROO","created_at":"2026-07-05T09:26:24.254338+00:00"},{"alias_kind":"pith_short_8","alias_value":"J3CY3Q7K","created_at":"2026-07-05T09:26:24.254338+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":3,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.30814","citing_title":"When Calibration Rankings Reverse: Accuracy-Controlled Evaluation for Fair Comparison of LLMs","ref_index":16,"is_internal_anchor":false},{"citing_arxiv_id":"2605.10202","citing_title":"Task-Aware Calibration: Provably Optimal Decoding in LLMs","ref_index":21,"is_internal_anchor":false},{"citing_arxiv_id":"2604.17200","citing_title":"Calibrating Model-Based Evaluation Metrics for Summarization","ref_index":62,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/J3CY3Q7K2Z3N7ROO7J7UNCLU3A","json":"https://pith.science/pith/J3CY3Q7K2Z3N7ROO7J7UNCLU3A.json","graph_json":"https://pith.science/api/pith-number/J3CY3Q7K2Z3N7ROO7J7UNCLU3A/graph.json","events_json":"https://pith.science/api/pith-number/J3CY3Q7K2Z3N7ROO7J7UNCLU3A/events.json","paper":"https://pith.science/paper/J3CY3Q7K"},"agent_actions":{"view_html":"https://pith.science/pith/J3CY3Q7K2Z3N7ROO7J7UNCLU3A","download_json":"https://pith.science/pith/J3CY3Q7K2Z3N7ROO7J7UNCLU3A.json","view_paper":"https://pith.science/paper/J3CY3Q7K","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2402.06544&json=true","fetch_graph":"https://pith.science/api/pith-number/J3CY3Q7K2Z3N7ROO7J7UNCLU3A/graph.json","fetch_events":"https://pith.science/api/pith-number/J3CY3Q7K2Z3N7ROO7J7UNCLU3A/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/J3CY3Q7K2Z3N7ROO7J7UNCLU3A/action/timestamp_anchor","attest_storage":"https://pith.science/pith/J3CY3Q7K2Z3N7ROO7J7UNCLU3A/action/storage_attestation","attest_author":"https://pith.science/pith/J3CY3Q7K2Z3N7ROO7J7UNCLU3A/action/author_attestation","sign_citation":"https://pith.science/pith/J3CY3Q7K2Z3N7ROO7J7UNCLU3A/action/citation_signature","submit_replication":"https://pith.science/pith/J3CY3Q7K2Z3N7ROO7J7UNCLU3A/action/replication_record"}},"created_at":"2026-07-05T09:26:24.254338+00:00","updated_at":"2026-07-05T09:26:24.254338+00:00"}