{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:SJPIGBBXX2FKEZD7YYJ3XZC6JH","short_pith_number":"pith:SJPIGBBX","schema_version":"1.0","canonical_sha256":"925e830437be8aa2647fc613bbe45e49fd4a95927b27b9990fa2d311e87cc98a","source":{"kind":"arxiv","id":"2310.01387","version":1},"attestation_state":"computed","paper":{"title":"It's MBR All the Way Down: Modern Generation Techniques Through the Lens of Minimum Bayes Risk","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Alex Xie, Amanda Bertsch, Graham Neubig, Matthew R. Gormley","submitted_at":"2023-10-02T17:47:10Z","abstract_excerpt":"Minimum Bayes Risk (MBR) decoding is a method for choosing the outputs of a machine learning system based not on the output with the highest probability, but the output with the lowest risk (expected error) among multiple candidates. It is a simple but powerful method: for an additional cost at inference time, MBR provides reliable several-point improvements across metrics for a wide variety of tasks without any additional data or training. Despite this, MBR is not frequently applied in NLP works, and knowledge of the method itself is limited. We first provide an introduction to the method and"},"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":"2310.01387","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2023-10-02T17:47:10Z","cross_cats_sorted":[],"title_canon_sha256":"338986f79c51999625971c7ded8cfbf527cb283fd8cded2bd3b5b975c444ff8b","abstract_canon_sha256":"1ef174f70a372995c19c3f2f3682edf397f58a756cb4c65f56c39f89103293f9"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T06:56:25.992501Z","signature_b64":"g6jekG2GkeyRXGBVO54S5jVVQBRxVb3gpb+5vAuAeer0v2BdOiFsXXldryiqIWiNJij9Eu0mbURUXb32KRXMAw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"925e830437be8aa2647fc613bbe45e49fd4a95927b27b9990fa2d311e87cc98a","last_reissued_at":"2026-07-05T06:56:25.991939Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T06:56:25.991939Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"It's MBR All the Way Down: Modern Generation Techniques Through the Lens of Minimum Bayes Risk","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Alex Xie, Amanda Bertsch, Graham Neubig, Matthew R. Gormley","submitted_at":"2023-10-02T17:47:10Z","abstract_excerpt":"Minimum Bayes Risk (MBR) decoding is a method for choosing the outputs of a machine learning system based not on the output with the highest probability, but the output with the lowest risk (expected error) among multiple candidates. It is a simple but powerful method: for an additional cost at inference time, MBR provides reliable several-point improvements across metrics for a wide variety of tasks without any additional data or training. Despite this, MBR is not frequently applied in NLP works, and knowledge of the method itself is limited. We first provide an introduction to the method and"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2310.01387","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/2310.01387/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":"2310.01387","created_at":"2026-07-05T06:56:25.991995+00:00"},{"alias_kind":"arxiv_version","alias_value":"2310.01387v1","created_at":"2026-07-05T06:56:25.991995+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2310.01387","created_at":"2026-07-05T06:56:25.991995+00:00"},{"alias_kind":"pith_short_12","alias_value":"SJPIGBBXX2FK","created_at":"2026-07-05T06:56:25.991995+00:00"},{"alias_kind":"pith_short_16","alias_value":"SJPIGBBXX2FKEZD7","created_at":"2026-07-05T06:56:25.991995+00:00"},{"alias_kind":"pith_short_8","alias_value":"SJPIGBBX","created_at":"2026-07-05T06:56:25.991995+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":1,"internal_anchor_count":1,"sample":[{"citing_arxiv_id":"2608.04001","citing_title":"Test-Time Scaling in Reasoning LLMs: Inference Regimes, Evaluation, and Reproducibility","ref_index":87,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/SJPIGBBXX2FKEZD7YYJ3XZC6JH","json":"https://pith.science/pith/SJPIGBBXX2FKEZD7YYJ3XZC6JH.json","graph_json":"https://pith.science/api/pith-number/SJPIGBBXX2FKEZD7YYJ3XZC6JH/graph.json","events_json":"https://pith.science/api/pith-number/SJPIGBBXX2FKEZD7YYJ3XZC6JH/events.json","paper":"https://pith.science/paper/SJPIGBBX"},"agent_actions":{"view_html":"https://pith.science/pith/SJPIGBBXX2FKEZD7YYJ3XZC6JH","download_json":"https://pith.science/pith/SJPIGBBXX2FKEZD7YYJ3XZC6JH.json","view_paper":"https://pith.science/paper/SJPIGBBX","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2310.01387&json=true","fetch_graph":"https://pith.science/api/pith-number/SJPIGBBXX2FKEZD7YYJ3XZC6JH/graph.json","fetch_events":"https://pith.science/api/pith-number/SJPIGBBXX2FKEZD7YYJ3XZC6JH/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/SJPIGBBXX2FKEZD7YYJ3XZC6JH/action/timestamp_anchor","attest_storage":"https://pith.science/pith/SJPIGBBXX2FKEZD7YYJ3XZC6JH/action/storage_attestation","attest_author":"https://pith.science/pith/SJPIGBBXX2FKEZD7YYJ3XZC6JH/action/author_attestation","sign_citation":"https://pith.science/pith/SJPIGBBXX2FKEZD7YYJ3XZC6JH/action/citation_signature","submit_replication":"https://pith.science/pith/SJPIGBBXX2FKEZD7YYJ3XZC6JH/action/replication_record"}},"created_at":"2026-07-05T06:56:25.991995+00:00","updated_at":"2026-07-05T06:56:25.991995+00:00"}