{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2023:XMILIN6C2UNAAKLQLQIRFLC5T5","short_pith_number":"pith:XMILIN6C","canonical_record":{"source":{"id":"2309.06657","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2023-09-13T01:07:25Z","cross_cats_sorted":[],"title_canon_sha256":"94c371d54ceabb0582b8bcfdf67f02e09b93b9e8698f1b426d516bc61f001cee","abstract_canon_sha256":"9dc78268a381bfb9e94936621d5d1120d76e61a58f43adc2e52b421185d705c9"},"schema_version":"1.0"},"canonical_sha256":"bb10b437c2d51a0029705c1112ac5d9f69b283307a69c0505faeec2e0e2aab36","source":{"kind":"arxiv","id":"2309.06657","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2309.06657","created_at":"2026-07-05T07:36:54Z"},{"alias_kind":"arxiv_version","alias_value":"2309.06657v2","created_at":"2026-07-05T07:36:54Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2309.06657","created_at":"2026-07-05T07:36:54Z"},{"alias_kind":"pith_short_12","alias_value":"XMILIN6C2UNA","created_at":"2026-07-05T07:36:54Z"},{"alias_kind":"pith_short_16","alias_value":"XMILIN6C2UNAAKLQ","created_at":"2026-07-05T07:36:54Z"},{"alias_kind":"pith_short_8","alias_value":"XMILIN6C","created_at":"2026-07-05T07:36:54Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2023:XMILIN6C2UNAAKLQLQIRFLC5T5","target":"record","payload":{"canonical_record":{"source":{"id":"2309.06657","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2023-09-13T01:07:25Z","cross_cats_sorted":[],"title_canon_sha256":"94c371d54ceabb0582b8bcfdf67f02e09b93b9e8698f1b426d516bc61f001cee","abstract_canon_sha256":"9dc78268a381bfb9e94936621d5d1120d76e61a58f43adc2e52b421185d705c9"},"schema_version":"1.0"},"canonical_sha256":"bb10b437c2d51a0029705c1112ac5d9f69b283307a69c0505faeec2e0e2aab36","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T07:36:54.159259Z","signature_b64":"AvRwHb1iv70KaMCe2skUgZfrCY9eSDPZ99HCxwiqABPSozMdo9vL/KzprPc0f5tx4fFUfGnW0rw+Y6dhSPhiBw==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"bb10b437c2d51a0029705c1112ac5d9f69b283307a69c0505faeec2e0e2aab36","last_reissued_at":"2026-07-05T07:36:54.158850Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T07:36:54.158850Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2309.06657","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":"2026-07-05T07:36:54Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"8oK5qGDvg1nMLoReRZdJ3PTA6TYh8Yvcdg4/kKWq0rC2fVh7T9HbsS2k7ABQ1rDvGKpm3i+/eitiny4ublbwCg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-07T12:31:39.401603Z"},"content_sha256":"604f738dabacb1e31d6f4a4c38410987c866a62acc82361ec65bee284f402b5c","schema_version":"1.0","event_id":"sha256:604f738dabacb1e31d6f4a4c38410987c866a62acc82361ec65bee284f402b5c"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2023:XMILIN6C2UNAAKLQLQIRFLC5T5","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Statistical Rejection Sampling Improves Preference Optimization","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.CL","authors_text":"Jialu Liu, Misha Khalman, Mohammad Saleh, Peter J. Liu, Rishabh Joshi, Tianqi Liu, Yao Zhao","submitted_at":"2023-09-13T01:07:25Z","abstract_excerpt":"Improving the alignment of language models with human preferences remains an active research challenge. Previous approaches have primarily utilized Reinforcement Learning from Human Feedback (RLHF) via online RL methods such as Proximal Policy Optimization (PPO). Recently, offline methods such as Sequence Likelihood Calibration (SLiC) and Direct Preference Optimization (DPO) have emerged as attractive alternatives, offering improvements in stability and scalability while maintaining competitive performance. SLiC refines its loss function using sequence pairs sampled from a supervised fine-tune"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2309.06657","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/2309.06657/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":"2026-07-05T07:36:54Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"eKC9v5Kt8YaD/9Uq8jv10+ZbzZ+tSWzA8SlPjLM4pb7Zn4Iuv0u6eb5hZZ8Pdvvcu85i/01UUK1dDF+NMXeFBA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-07T12:31:39.401967Z"},"content_sha256":"61e8e5afe369dc9336c03bf991122afbcb0847d4c3eb096f1f4548b312033f6c","schema_version":"1.0","event_id":"sha256:61e8e5afe369dc9336c03bf991122afbcb0847d4c3eb096f1f4548b312033f6c"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/XMILIN6C2UNAAKLQLQIRFLC5T5/bundle.json","state_url":"https://pith.science/pith/XMILIN6C2UNAAKLQLQIRFLC5T5/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/XMILIN6C2UNAAKLQLQIRFLC5T5/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":"2026-07-07T12:31:39Z","links":{"resolver":"https://pith.science/pith/XMILIN6C2UNAAKLQLQIRFLC5T5","bundle":"https://pith.science/pith/XMILIN6C2UNAAKLQLQIRFLC5T5/bundle.json","state":"https://pith.science/pith/XMILIN6C2UNAAKLQLQIRFLC5T5/state.json","well_known_bundle":"https://pith.science/.well-known/pith/XMILIN6C2UNAAKLQLQIRFLC5T5/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2023:XMILIN6C2UNAAKLQLQIRFLC5T5","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":"9dc78268a381bfb9e94936621d5d1120d76e61a58f43adc2e52b421185d705c9","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2023-09-13T01:07:25Z","title_canon_sha256":"94c371d54ceabb0582b8bcfdf67f02e09b93b9e8698f1b426d516bc61f001cee"},"schema_version":"1.0","source":{"id":"2309.06657","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2309.06657","created_at":"2026-07-05T07:36:54Z"},{"alias_kind":"arxiv_version","alias_value":"2309.06657v2","created_at":"2026-07-05T07:36:54Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2309.06657","created_at":"2026-07-05T07:36:54Z"},{"alias_kind":"pith_short_12","alias_value":"XMILIN6C2UNA","created_at":"2026-07-05T07:36:54Z"},{"alias_kind":"pith_short_16","alias_value":"XMILIN6C2UNAAKLQ","created_at":"2026-07-05T07:36:54Z"},{"alias_kind":"pith_short_8","alias_value":"XMILIN6C","created_at":"2026-07-05T07:36:54Z"}],"graph_snapshots":[{"event_id":"sha256:61e8e5afe369dc9336c03bf991122afbcb0847d4c3eb096f1f4548b312033f6c","target":"graph","created_at":"2026-07-05T07:36:54Z","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/2309.06657/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Improving the alignment of language models with human preferences remains an active research challenge. Previous approaches have primarily utilized Reinforcement Learning from Human Feedback (RLHF) via online RL methods such as Proximal Policy Optimization (PPO). Recently, offline methods such as Sequence Likelihood Calibration (SLiC) and Direct Preference Optimization (DPO) have emerged as attractive alternatives, offering improvements in stability and scalability while maintaining competitive performance. SLiC refines its loss function using sequence pairs sampled from a supervised fine-tune","authors_text":"Jialu Liu, Misha Khalman, Mohammad Saleh, Peter J. Liu, Rishabh Joshi, Tianqi Liu, Yao Zhao","cross_cats":[],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2023-09-13T01:07:25Z","title":"Statistical Rejection Sampling Improves Preference Optimization"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2309.06657","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":"sha256:604f738dabacb1e31d6f4a4c38410987c866a62acc82361ec65bee284f402b5c","target":"record","created_at":"2026-07-05T07:36:54Z","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":"9dc78268a381bfb9e94936621d5d1120d76e61a58f43adc2e52b421185d705c9","cross_cats_sorted":[],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.CL","submitted_at":"2023-09-13T01:07:25Z","title_canon_sha256":"94c371d54ceabb0582b8bcfdf67f02e09b93b9e8698f1b426d516bc61f001cee"},"schema_version":"1.0","source":{"id":"2309.06657","kind":"arxiv","version":2}},"canonical_sha256":"bb10b437c2d51a0029705c1112ac5d9f69b283307a69c0505faeec2e0e2aab36","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"bb10b437c2d51a0029705c1112ac5d9f69b283307a69c0505faeec2e0e2aab36","first_computed_at":"2026-07-05T07:36:54.158850Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T07:36:54.158850Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"AvRwHb1iv70KaMCe2skUgZfrCY9eSDPZ99HCxwiqABPSozMdo9vL/KzprPc0f5tx4fFUfGnW0rw+Y6dhSPhiBw==","signature_status":"signed_v1","signed_at":"2026-07-05T07:36:54.159259Z","signed_message":"canonical_sha256_bytes"},"source_id":"2309.06657","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:604f738dabacb1e31d6f4a4c38410987c866a62acc82361ec65bee284f402b5c","sha256:61e8e5afe369dc9336c03bf991122afbcb0847d4c3eb096f1f4548b312033f6c"],"state_sha256":"935d8252b915e9836feef68ec150d58501c7c2b569889452a8c77a2fbc3a24a1"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"C56698iwQc92AQtW0FXfvz7t27H5eMHVuw7Qa0/SYNXv07rYwKmpPFMoQRmdDsr8R8hjDYGu0t8L92RGV0PnBg==","signed_message":"bundle_sha256_bytes","signed_at":"2026-07-07T12:31:39.403866Z","bundle_sha256":"be6d7c39b94b23fb46780a4f40bea4f0ded2267a82e8816eda0927bf5e883db1"}}