{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2023:GAWBHK2KM63L5JGWFRCVV3YOXY","short_pith_number":"pith:GAWBHK2K","canonical_record":{"source":{"id":"2311.08380","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2023-11-14T18:43:51Z","cross_cats_sorted":[],"title_canon_sha256":"fb1ea103183e383024dabe61eb3f4a5c7e4b6ef4d2cecdad61f404b550479960","abstract_canon_sha256":"72f453712993833213946a59a565e96014bd470e5730e4f68169b2a7af363ba6"},"schema_version":"1.0"},"canonical_sha256":"302c13ab4a67b6bea4d62c455aef0ebe0b309eb2e7c6ee77417adc08c1e12652","source":{"kind":"arxiv","id":"2311.08380","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2311.08380","created_at":"2026-07-05T08:07:12Z"},{"alias_kind":"arxiv_version","alias_value":"2311.08380v2","created_at":"2026-07-05T08:07:12Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2311.08380","created_at":"2026-07-05T08:07:12Z"},{"alias_kind":"pith_short_12","alias_value":"GAWBHK2KM63L","created_at":"2026-07-05T08:07:12Z"},{"alias_kind":"pith_short_16","alias_value":"GAWBHK2KM63L5JGW","created_at":"2026-07-05T08:07:12Z"},{"alias_kind":"pith_short_8","alias_value":"GAWBHK2K","created_at":"2026-07-05T08:07:12Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2023:GAWBHK2KM63L5JGWFRCVV3YOXY","target":"record","payload":{"canonical_record":{"source":{"id":"2311.08380","kind":"arxiv","version":2},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2023-11-14T18:43:51Z","cross_cats_sorted":[],"title_canon_sha256":"fb1ea103183e383024dabe61eb3f4a5c7e4b6ef4d2cecdad61f404b550479960","abstract_canon_sha256":"72f453712993833213946a59a565e96014bd470e5730e4f68169b2a7af363ba6"},"schema_version":"1.0"},"canonical_sha256":"302c13ab4a67b6bea4d62c455aef0ebe0b309eb2e7c6ee77417adc08c1e12652","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:07:12.054797Z","signature_b64":"AOUPOk8yC4A2k7cbZDHJA/3SyPAixWSOQaWZ6tFsv6tTNby4XzVAa1qJQBLrhqLjLf5PDWxVaPjZ477Plb7ACQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"302c13ab4a67b6bea4d62c455aef0ebe0b309eb2e7c6ee77417adc08c1e12652","last_reissued_at":"2026-07-05T08:07:12.054276Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:07:12.054276Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2311.08380","source_version":2,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T08:07:12Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"tkpQXoObxKQ6p9DdyF8ChGyD//v7kA/4ya3C6MpIadvFYoS4gx6IzxPMCGlAUYAx0beQZikIE+vFVMuk5Z35Bw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-13T19:55:56.907438Z"},"content_sha256":"03c07ec3ba78d0c2503baaba690b2ec0bac7e84d737a2b930a3ced4509504c15","schema_version":"1.0","event_id":"sha256:03c07ec3ba78d0c2503baaba690b2ec0bac7e84d737a2b930a3ced4509504c15"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2023:GAWBHK2KM63L5JGWFRCVV3YOXY","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Direct Preference Optimization for Neural Machine Translation with Minimum Bayes Risk Decoding","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Bill Byrne, Guangyu Yang, Jinghong Chen, Weizhe Lin","submitted_at":"2023-11-14T18:43:51Z","abstract_excerpt":"Minimum Bayes Risk (MBR) decoding can significantly improve translation performance of Multilingual Large Language Models (MLLMs). However, MBR decoding is computationally expensive. We show how the recently developed Reinforcement Learning technique, Direct Preference Optimization (DPO), can fine-tune MLLMs to get the gains of MBR without any additional computation in inference. Our method uses only a small monolingual fine-tuning set and yields significantly improved performance on multiple NMT test sets compared to MLLMs without DPO."},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2311.08380","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/2311.08380/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"},"verdict_id":null},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T08:07:12Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"67HBimagi1T7OmIqo8YyPOdk3zmZ/Pb8Q63m1Jf/zbE5PymusCcGmxNiVq30cCHKq0Cnc7zKanql/Kgm4YqHAw==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-13T19:55:56.909589Z"},"content_sha256":"8f847358dfd06f0d0917f9b7ad8d863c13d8673f92c13678554eb5c14bfe9050","schema_version":"1.0","event_id":"sha256:8f847358dfd06f0d0917f9b7ad8d863c13d8673f92c13678554eb5c14bfe9050"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/GAWBHK2KM63L5JGWFRCVV3YOXY/bundle.json","state_url":"https://pith.science/pith/GAWBHK2KM63L5JGWFRCVV3YOXY/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/GAWBHK2KM63L5JGWFRCVV3YOXY/bundle.json","status":"primary"}],"public_keys":[{"key_id":"pith-v1-2026-05","algorithm":"ed25519","format":"raw","public_key_b64":"stVStoiQhXFxp4s2pdzPNoqVNBMojDU/fJ2db5S3CbM=","public_key_hex":"b2d552b68890857171a78b36a5dccf368a953413288c353f7c9d9d6f94b709b3","fingerprint_sha256_b32_first128bits":"RVFV5Z2OI2J3ZUO7ERDEBCYNKS","fingerprint_sha256_hex":"8d4b5ee74e4693bcd1df2446408b0d54","rotates_at":null,"url":"https://pith.science/pith-signing-key.json","notes":"Pith uses this Ed25519 key to sign canonical record SHA-256 digests. Verify with: ed25519_verify(public_key, message=canonical_sha256_bytes, signature=base64decode(signature_b64))."}],"merge_version":"pith-open-graph-merge-v1","built_at":"2026-08-13T19:55:56Z","links":{"resolver":"https://pith.science/pith/GAWBHK2KM63L5JGWFRCVV3YOXY","bundle":"https://pith.science/pith/GAWBHK2KM63L5JGWFRCVV3YOXY/bundle.json","state":"https://pith.science/pith/GAWBHK2KM63L5JGWFRCVV3YOXY/state.json","well_known_bundle":"https://pith.science/.well-known/pith/GAWBHK2KM63L5JGWFRCVV3YOXY/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2023:GAWBHK2KM63L5JGWFRCVV3YOXY","merge_version":"pith-open-graph-merge-v1","event_count":2,"valid_event_count":2,"invalid_event_count":0,"equivocation_count":0,"current":{"canonical_record":{"metadata":{"abstract_canon_sha256":"72f453712993833213946a59a565e96014bd470e5730e4f68169b2a7af363ba6","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2023-11-14T18:43:51Z","title_canon_sha256":"fb1ea103183e383024dabe61eb3f4a5c7e4b6ef4d2cecdad61f404b550479960"},"schema_version":"1.0","source":{"id":"2311.08380","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2311.08380","created_at":"2026-07-05T08:07:12Z"},{"alias_kind":"arxiv_version","alias_value":"2311.08380v2","created_at":"2026-07-05T08:07:12Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2311.08380","created_at":"2026-07-05T08:07:12Z"},{"alias_kind":"pith_short_12","alias_value":"GAWBHK2KM63L","created_at":"2026-07-05T08:07:12Z"},{"alias_kind":"pith_short_16","alias_value":"GAWBHK2KM63L5JGW","created_at":"2026-07-05T08:07:12Z"},{"alias_kind":"pith_short_8","alias_value":"GAWBHK2K","created_at":"2026-07-05T08:07:12Z"}],"graph_snapshots":[{"event_id":"sha256:8f847358dfd06f0d0917f9b7ad8d863c13d8673f92c13678554eb5c14bfe9050","target":"graph","created_at":"2026-07-05T08:07:12Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"graph_snapshot":{"author_claims":{"count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57","strong_count":0},"builder_version":"pith-number-builder-2026-05-17-v1","claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"formal_canon":{"evidence_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"integrity":{"available":true,"clean":true,"detectors_run":[],"endpoint":"/pith/2311.08380/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Minimum Bayes Risk (MBR) decoding can significantly improve translation performance of Multilingual Large Language Models (MLLMs). However, MBR decoding is computationally expensive. We show how the recently developed Reinforcement Learning technique, Direct Preference Optimization (DPO), can fine-tune MLLMs to get the gains of MBR without any additional computation in inference. Our method uses only a small monolingual fine-tuning set and yields significantly improved performance on multiple NMT test sets compared to MLLMs without DPO.","authors_text":"Bill Byrne, Guangyu Yang, Jinghong Chen, Weizhe Lin","cross_cats":[],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2023-11-14T18:43:51Z","title":"Direct Preference Optimization for Neural Machine Translation with Minimum Bayes Risk Decoding"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2311.08380","kind":"arxiv","version":2},"verdict":{"created_at":null,"id":null,"model_set":{},"one_line_summary":"","pipeline_version":null,"pith_extraction_headline":"","strongest_claim":"","weakest_assumption":""}},"verdict_id":null}}],"author_attestations":[],"timestamp_anchors":[],"storage_attestations":[],"citation_signatures":[],"replication_records":[],"corrections":[],"mirror_hints":[],"record_created":{"event_id":"sha256:03c07ec3ba78d0c2503baaba690b2ec0bac7e84d737a2b930a3ced4509504c15","target":"record","created_at":"2026-07-05T08:07:12Z","signer":{"key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signer_id":"pith.science","signer_type":"pith_registry"},"payload":{"attestation_state":"computed","canonical_record":{"metadata":{"abstract_canon_sha256":"72f453712993833213946a59a565e96014bd470e5730e4f68169b2a7af363ba6","cross_cats_sorted":[],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2023-11-14T18:43:51Z","title_canon_sha256":"fb1ea103183e383024dabe61eb3f4a5c7e4b6ef4d2cecdad61f404b550479960"},"schema_version":"1.0","source":{"id":"2311.08380","kind":"arxiv","version":2}},"canonical_sha256":"302c13ab4a67b6bea4d62c455aef0ebe0b309eb2e7c6ee77417adc08c1e12652","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"302c13ab4a67b6bea4d62c455aef0ebe0b309eb2e7c6ee77417adc08c1e12652","first_computed_at":"2026-07-05T08:07:12.054276Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T08:07:12.054276Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"AOUPOk8yC4A2k7cbZDHJA/3SyPAixWSOQaWZ6tFsv6tTNby4XzVAa1qJQBLrhqLjLf5PDWxVaPjZ477Plb7ACQ==","signature_status":"signed_v1","signed_at":"2026-07-05T08:07:12.054797Z","signed_message":"canonical_sha256_bytes"},"source_id":"2311.08380","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:03c07ec3ba78d0c2503baaba690b2ec0bac7e84d737a2b930a3ced4509504c15","sha256:8f847358dfd06f0d0917f9b7ad8d863c13d8673f92c13678554eb5c14bfe9050"],"state_sha256":"2a7003e495f5a9ed7ad7ce639af1defdb04ab9c9a2b5d1d405aab5102006310f"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"MBBnxlGvXtsyJgNGyELdfOqLpM5ZehPG1A2SkVOozdzWs/3+ULwF79zgo7rSdybiOMPBbI+kNcUJmY3MpQebCw==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-13T19:55:56.940444Z","bundle_sha256":"8d5f183ff511d1179edf9a8e1a78a5b07fb476cfd1228a6e972964dd955941c0"}}