{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2024:YOITWZFAFAGGM4E6OUH2DJQNTR","short_pith_number":"pith:YOITWZFA","canonical_record":{"source":{"id":"2403.19279","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-03-28T10:02:10Z","cross_cats_sorted":["cs.AI","cs.LG"],"title_canon_sha256":"8e288d820f10d63806ffe15dea45d8aca2a129ff52aff9d9ea6a80318fda4a49","abstract_canon_sha256":"a4105cd95da7a91669deebe7979dedefad772a98cc67d54920c3e95c79cea7d1"},"schema_version":"1.0"},"canonical_sha256":"c3913b64a0280c66709e750fa1a60d9c6651321ebd9740e9944ebf110793a673","source":{"kind":"arxiv","id":"2403.19279","version":1},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2403.19279","created_at":"2026-07-05T08:01:47Z"},{"alias_kind":"arxiv_version","alias_value":"2403.19279v1","created_at":"2026-07-05T08:01:47Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2403.19279","created_at":"2026-07-05T08:01:47Z"},{"alias_kind":"pith_short_12","alias_value":"YOITWZFAFAGG","created_at":"2026-07-05T08:01:47Z"},{"alias_kind":"pith_short_16","alias_value":"YOITWZFAFAGGM4E6","created_at":"2026-07-05T08:01:47Z"},{"alias_kind":"pith_short_8","alias_value":"YOITWZFA","created_at":"2026-07-05T08:01:47Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2024:YOITWZFAFAGGM4E6OUH2DJQNTR","target":"record","payload":{"canonical_record":{"source":{"id":"2403.19279","kind":"arxiv","version":1},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-03-28T10:02:10Z","cross_cats_sorted":["cs.AI","cs.LG"],"title_canon_sha256":"8e288d820f10d63806ffe15dea45d8aca2a129ff52aff9d9ea6a80318fda4a49","abstract_canon_sha256":"a4105cd95da7a91669deebe7979dedefad772a98cc67d54920c3e95c79cea7d1"},"schema_version":"1.0"},"canonical_sha256":"c3913b64a0280c66709e750fa1a60d9c6651321ebd9740e9944ebf110793a673","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:01:47.235024Z","signature_b64":"EGOYTeaT15I0J155Fye4eUe8VNXaxwpwEuoO7Kkd56sLkfsOJWlTgYN17HFFm/po527lyNYNg5Q/C0CZFNTBAg==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"c3913b64a0280c66709e750fa1a60d9c6651321ebd9740e9944ebf110793a673","last_reissued_at":"2026-07-05T08:01:47.234648Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:01:47.234648Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2403.19279","source_version":1,"attestation_state":"computed"},"signer":{"signer_id":"pith.science","signer_type":"pith_registry","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"created_at":"2026-07-05T08:01:47Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"5CUO2iFiRxFWNLn+K6Z1nfsLZ7DtLr/u/E90RYVbHbhvDrG6atJR1H+fx6vFHIG330+Bs6u6afZQIimJi3EyBg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-14T07:06:50.447635Z"},"content_sha256":"c41d8301bc2471e4ae29c9f5cd53e2083864412ce9f0d4e97d6c71c60ecaaad0","schema_version":"1.0","event_id":"sha256:c41d8301bc2471e4ae29c9f5cd53e2083864412ce9f0d4e97d6c71c60ecaaad0"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2024:YOITWZFAFAGGM4E6OUH2DJQNTR","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Fine-Tuning Language Models with Reward Learning on Policy","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"","cross_cats":["cs.AI","cs.LG"],"primary_cat":"cs.CL","authors_text":"Fei Huang, Hao Lang, Yongbin Li","submitted_at":"2024-03-28T10:02:10Z","abstract_excerpt":"Reinforcement learning from human feedback (RLHF) has emerged as an effective approach to aligning large language models (LLMs) to human preferences. RLHF contains three steps, i.e., human preference collecting, reward learning, and policy optimization, which are usually performed serially. Despite its popularity, however, (fixed) reward models may suffer from inaccurate off-distribution, since policy optimization continuously shifts LLMs' data distribution. Repeatedly collecting new preference data from the latest LLMs may alleviate this issue, which unfortunately makes the resulting system m"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2403.19279","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/2403.19279/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":"2026-07-05T08:01:47Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"tIVXZ80VOYf+CARFiMkVFOZg8WYYIN2h76yvH8W9v8yivnAjwn9nQBgx1g+HnO6QzpI/inxoQo8FLruEARXbDA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-08-14T07:06:50.448203Z"},"content_sha256":"adda03d1d1f6b4039fe917525222192e4f3055bed3cc14d800d82a71237ebfdb","schema_version":"1.0","event_id":"sha256:adda03d1d1f6b4039fe917525222192e4f3055bed3cc14d800d82a71237ebfdb"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/YOITWZFAFAGGM4E6OUH2DJQNTR/bundle.json","state_url":"https://pith.science/pith/YOITWZFAFAGGM4E6OUH2DJQNTR/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/YOITWZFAFAGGM4E6OUH2DJQNTR/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":"2026-08-14T07:06:50Z","links":{"resolver":"https://pith.science/pith/YOITWZFAFAGGM4E6OUH2DJQNTR","bundle":"https://pith.science/pith/YOITWZFAFAGGM4E6OUH2DJQNTR/bundle.json","state":"https://pith.science/pith/YOITWZFAFAGGM4E6OUH2DJQNTR/state.json","well_known_bundle":"https://pith.science/.well-known/pith/YOITWZFAFAGGM4E6OUH2DJQNTR/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:YOITWZFAFAGGM4E6OUH2DJQNTR","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":"a4105cd95da7a91669deebe7979dedefad772a98cc67d54920c3e95c79cea7d1","cross_cats_sorted":["cs.AI","cs.LG"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-03-28T10:02:10Z","title_canon_sha256":"8e288d820f10d63806ffe15dea45d8aca2a129ff52aff9d9ea6a80318fda4a49"},"schema_version":"1.0","source":{"id":"2403.19279","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2403.19279","created_at":"2026-07-05T08:01:47Z"},{"alias_kind":"arxiv_version","alias_value":"2403.19279v1","created_at":"2026-07-05T08:01:47Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2403.19279","created_at":"2026-07-05T08:01:47Z"},{"alias_kind":"pith_short_12","alias_value":"YOITWZFAFAGG","created_at":"2026-07-05T08:01:47Z"},{"alias_kind":"pith_short_16","alias_value":"YOITWZFAFAGGM4E6","created_at":"2026-07-05T08:01:47Z"},{"alias_kind":"pith_short_8","alias_value":"YOITWZFA","created_at":"2026-07-05T08:01:47Z"}],"graph_snapshots":[{"event_id":"sha256:adda03d1d1f6b4039fe917525222192e4f3055bed3cc14d800d82a71237ebfdb","target":"graph","created_at":"2026-07-05T08:01:47Z","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/2403.19279/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Reinforcement learning from human feedback (RLHF) has emerged as an effective approach to aligning large language models (LLMs) to human preferences. RLHF contains three steps, i.e., human preference collecting, reward learning, and policy optimization, which are usually performed serially. Despite its popularity, however, (fixed) reward models may suffer from inaccurate off-distribution, since policy optimization continuously shifts LLMs' data distribution. Repeatedly collecting new preference data from the latest LLMs may alleviate this issue, which unfortunately makes the resulting system m","authors_text":"Fei Huang, Hao Lang, Yongbin Li","cross_cats":["cs.AI","cs.LG"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-03-28T10:02:10Z","title":"Fine-Tuning Language Models with Reward Learning on Policy"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2403.19279","kind":"arxiv","version":1},"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:c41d8301bc2471e4ae29c9f5cd53e2083864412ce9f0d4e97d6c71c60ecaaad0","target":"record","created_at":"2026-07-05T08:01:47Z","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":"a4105cd95da7a91669deebe7979dedefad772a98cc67d54920c3e95c79cea7d1","cross_cats_sorted":["cs.AI","cs.LG"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-03-28T10:02:10Z","title_canon_sha256":"8e288d820f10d63806ffe15dea45d8aca2a129ff52aff9d9ea6a80318fda4a49"},"schema_version":"1.0","source":{"id":"2403.19279","kind":"arxiv","version":1}},"canonical_sha256":"c3913b64a0280c66709e750fa1a60d9c6651321ebd9740e9944ebf110793a673","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"c3913b64a0280c66709e750fa1a60d9c6651321ebd9740e9944ebf110793a673","first_computed_at":"2026-07-05T08:01:47.234648Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T08:01:47.234648Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"EGOYTeaT15I0J155Fye4eUe8VNXaxwpwEuoO7Kkd56sLkfsOJWlTgYN17HFFm/po527lyNYNg5Q/C0CZFNTBAg==","signature_status":"signed_v1","signed_at":"2026-07-05T08:01:47.235024Z","signed_message":"canonical_sha256_bytes"},"source_id":"2403.19279","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:c41d8301bc2471e4ae29c9f5cd53e2083864412ce9f0d4e97d6c71c60ecaaad0","sha256:adda03d1d1f6b4039fe917525222192e4f3055bed3cc14d800d82a71237ebfdb"],"state_sha256":"065421614bc54eec854f9cdcaed0f1a0ebe3413eaa80d43e10a3de4931021fdd"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"ZYmlkvG0TdmgspMDNFqLY/mA1LeoHCwXfsVt8dlCHu+6yu0pOQQpkYZNe4hqVOHcE9SFqsfElziaWmWeuQYGCg==","signed_message":"bundle_sha256_bytes","signed_at":"2026-08-14T07:06:50.453195Z","bundle_sha256":"9c084286149196b21e15a8ba63a4117197e455928c937cb32afdad7f41ac0cf5"}}