{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:PVGXLFGYQPO4ABUOWT4XYTKXSN","short_pith_number":"pith:PVGXLFGY","schema_version":"1.0","canonical_sha256":"7d4d7594d883ddc0068eb4f97c4d579376f738e68037347c223b52a4b7d1557d","source":{"kind":"arxiv","id":"2410.02902","version":4},"attestation_state":"computed","paper":{"title":"Better Instruction-Following Through Minimum Bayes Risk","license":"http://creativecommons.org/licenses/by-sa/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Amanda Bertsch, Graham Neubig, Ian Wu, Patrick Fernandes, Seungone Kim, Sina Pakazad","submitted_at":"2024-10-03T18:48:38Z","abstract_excerpt":"General-purpose LLM judges capable of human-level evaluation provide not only a scalable and accurate way of evaluating instruction-following LLMs but also new avenues for supervising and improving their performance. One promising way of leveraging LLM judges for supervision is through Minimum Bayes Risk (MBR) decoding, which uses a reference-based evaluator to select a high-quality output from amongst a set of candidate outputs. In the first part of this work, we explore using MBR decoding as a method for improving the test-time performance of instruction-following LLMs. We find that MBR deco"},"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":"2410.02902","kind":"arxiv","version":4},"metadata":{"license":"http://creativecommons.org/licenses/by-sa/4.0/","primary_cat":"cs.CL","submitted_at":"2024-10-03T18:48:38Z","cross_cats_sorted":["cs.AI"],"title_canon_sha256":"6c465bcb02f8184bce92523e2ee6de6177efcf3d30bdd2b794b9ed219ef969fe","abstract_canon_sha256":"1e2ab7d11667ebcafd23e4c4df8171135c31701c2c36404db5fd95a8e6cf4ac4"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:19:56.007915Z","signature_b64":"+IoNbsRw72D24+FANE6kVam/cvSJDoAUr1NC0Pja/mUHC5aJHr950IFG3iGVr3kgbu9ClOycbC42S258JeQxCQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"7d4d7594d883ddc0068eb4f97c4d579376f738e68037347c223b52a4b7d1557d","last_reissued_at":"2026-07-05T10:19:56.007441Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:19:56.007441Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"Better Instruction-Following Through Minimum Bayes Risk","license":"http://creativecommons.org/licenses/by-sa/4.0/","headline":"","cross_cats":["cs.AI"],"primary_cat":"cs.CL","authors_text":"Amanda Bertsch, Graham Neubig, Ian Wu, Patrick Fernandes, Seungone Kim, Sina Pakazad","submitted_at":"2024-10-03T18:48:38Z","abstract_excerpt":"General-purpose LLM judges capable of human-level evaluation provide not only a scalable and accurate way of evaluating instruction-following LLMs but also new avenues for supervising and improving their performance. One promising way of leveraging LLM judges for supervision is through Minimum Bayes Risk (MBR) decoding, which uses a reference-based evaluator to select a high-quality output from amongst a set of candidate outputs. In the first part of this work, we explore using MBR decoding as a method for improving the test-time performance of instruction-following LLMs. We find that MBR deco"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2410.02902","kind":"arxiv","version":4},"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/2410.02902/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":"2410.02902","created_at":"2026-07-05T10:19:56.007498+00:00"},{"alias_kind":"arxiv_version","alias_value":"2410.02902v4","created_at":"2026-07-05T10:19:56.007498+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2410.02902","created_at":"2026-07-05T10:19:56.007498+00:00"},{"alias_kind":"pith_short_12","alias_value":"PVGXLFGYQPO4","created_at":"2026-07-05T10:19:56.007498+00:00"},{"alias_kind":"pith_short_16","alias_value":"PVGXLFGYQPO4ABUO","created_at":"2026-07-05T10:19:56.007498+00:00"},{"alias_kind":"pith_short_8","alias_value":"PVGXLFGY","created_at":"2026-07-05T10:19:56.007498+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2412.21187","citing_title":"Do NOT Think That Much for 2+3=? On the Overthinking of o1-Like LLMs","ref_index":284,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/PVGXLFGYQPO4ABUOWT4XYTKXSN","json":"https://pith.science/pith/PVGXLFGYQPO4ABUOWT4XYTKXSN.json","graph_json":"https://pith.science/api/pith-number/PVGXLFGYQPO4ABUOWT4XYTKXSN/graph.json","events_json":"https://pith.science/api/pith-number/PVGXLFGYQPO4ABUOWT4XYTKXSN/events.json","paper":"https://pith.science/paper/PVGXLFGY"},"agent_actions":{"view_html":"https://pith.science/pith/PVGXLFGYQPO4ABUOWT4XYTKXSN","download_json":"https://pith.science/pith/PVGXLFGYQPO4ABUOWT4XYTKXSN.json","view_paper":"https://pith.science/paper/PVGXLFGY","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2410.02902&json=true","fetch_graph":"https://pith.science/api/pith-number/PVGXLFGYQPO4ABUOWT4XYTKXSN/graph.json","fetch_events":"https://pith.science/api/pith-number/PVGXLFGYQPO4ABUOWT4XYTKXSN/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/PVGXLFGYQPO4ABUOWT4XYTKXSN/action/timestamp_anchor","attest_storage":"https://pith.science/pith/PVGXLFGYQPO4ABUOWT4XYTKXSN/action/storage_attestation","attest_author":"https://pith.science/pith/PVGXLFGYQPO4ABUOWT4XYTKXSN/action/author_attestation","sign_citation":"https://pith.science/pith/PVGXLFGYQPO4ABUOWT4XYTKXSN/action/citation_signature","submit_replication":"https://pith.science/pith/PVGXLFGYQPO4ABUOWT4XYTKXSN/action/replication_record"}},"created_at":"2026-07-05T10:19:56.007498+00:00","updated_at":"2026-07-05T10:19:56.007498+00:00"}