{"bundle_type":"pith_open_graph_bundle","bundle_version":"1.0","pith_number":"pith:2025:6SRJ2EZQPXK4MVPFGEOE5YN3KE","short_pith_number":"pith:6SRJ2EZQ","canonical_record":{"source":{"id":"2502.00203","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-01-31T22:39:04Z","cross_cats_sorted":["cs.CL"],"title_canon_sha256":"2689c97482820d5c2bd22a10a9951448b9bdb819dc5026599e021ed9023cceb0","abstract_canon_sha256":"6010c76888d9fae9c50a284ade283c029ae3454cf3e10c7bbee74b44ef551dff"},"schema_version":"1.0"},"canonical_sha256":"f4a29d13307dd5c655e5311c4ee1bb51127e338dcdbd8353a0ae7274305984a1","source":{"kind":"arxiv","id":"2502.00203","version":2},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2502.00203","created_at":"2026-07-05T10:11:32Z"},{"alias_kind":"arxiv_version","alias_value":"2502.00203v2","created_at":"2026-07-05T10:11:32Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2502.00203","created_at":"2026-07-05T10:11:32Z"},{"alias_kind":"pith_short_12","alias_value":"6SRJ2EZQPXK4","created_at":"2026-07-05T10:11:32Z"},{"alias_kind":"pith_short_16","alias_value":"6SRJ2EZQPXK4MVPF","created_at":"2026-07-05T10:11:32Z"},{"alias_kind":"pith_short_8","alias_value":"6SRJ2EZQ","created_at":"2026-07-05T10:11:32Z"}],"events":[{"event_type":"record_created","subject_pith_number":"pith:2025:6SRJ2EZQPXK4MVPFGEOE5YN3KE","target":"record","payload":{"canonical_record":{"source":{"id":"2502.00203","kind":"arxiv","version":2},"metadata":{"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-01-31T22:39:04Z","cross_cats_sorted":["cs.CL"],"title_canon_sha256":"2689c97482820d5c2bd22a10a9951448b9bdb819dc5026599e021ed9023cceb0","abstract_canon_sha256":"6010c76888d9fae9c50a284ade283c029ae3454cf3e10c7bbee74b44ef551dff"},"schema_version":"1.0"},"canonical_sha256":"f4a29d13307dd5c655e5311c4ee1bb51127e338dcdbd8353a0ae7274305984a1","receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T10:11:32.687141Z","signature_b64":"Qz8btzo9q2+zHpKRJw617f4kTT5e5DeFzmZH1dtWeZnzOKfR0bVYOXXUM6P6KxLmZeq+Xa51iqhQDn4S67f0CQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"f4a29d13307dd5c655e5311c4ee1bb51127e338dcdbd8353a0ae7274305984a1","last_reissued_at":"2026-07-05T10:11:32.686495Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T10:11:32.686495Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"source_kind":"arxiv","source_id":"2502.00203","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":"2026-07-05T10:11:32Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"mjn3AWvddlzpQfLGlkulrDLw772WbMXoS3cyE1+OUt2FoP6fyQPmDynpo1CXgCeUCCJz+FQPz+6llTawVUFeDA==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-07T13:48:37.593710Z"},"content_sha256":"00ac48166d937f6e423e5e97947e8762ae3d61818d283015fe7230e9d2d23f04","schema_version":"1.0","event_id":"sha256:00ac48166d937f6e423e5e97947e8762ae3d61818d283015fe7230e9d2d23f04"},{"event_type":"graph_snapshot","subject_pith_number":"pith:2025:6SRJ2EZQPXK4MVPFGEOE5YN3KE","target":"graph","payload":{"graph_snapshot":{"paper":{"title":"Reward-aware Preference Optimization: A Unified Mathematical Framework for Model Alignment","license":"http://creativecommons.org/licenses/by/4.0/","headline":"","cross_cats":["cs.CL"],"primary_cat":"cs.LG","authors_text":"Adithya Renduchintala, Alexander Bukharin, David Mosallanezhad, Dmitry Chichkov, Gerald Shen, Jiaqi Zeng, Oleksii Kuchaiev, Olivier Delalleau, Shengyang Sun, Soumye Singhal, Tugrul Konuk, Yian Zhang, Yi Dong, Zhilin Wang","submitted_at":"2025-01-31T22:39:04Z","abstract_excerpt":"The rapid development of large language model (LLM) alignment algorithms has resulted in a complex and fragmented landscape, with limited clarity on the effectiveness of different methods and their inter-connections. This paper introduces Reward-Aware Preference Optimization (RPO), a mathematical framework that unifies popular preference optimization techniques in LLM alignment, including DPO, IPO, SimPO, and REINFORCE (LOO), among others. RPO provides a structured approach to disentangle and systematically study the impact of various design choices, such as the optimization objective, the num"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2502.00203","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/2502.00203/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":"2026-07-05T10:11:32Z","supersedes":[],"prev_event":null,"signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"iKrx3MKwdF4dNePKqnsXkgh1xYz59UsCf6EFKyrcN+c89gXfUFEqM6aLbe+Cn9s+Mo0OsLbMQqoKS2iXWW8fAg==","signed_message":"open_graph_event_sha256_bytes","signed_at":"2026-07-07T13:48:37.594091Z"},"content_sha256":"55547ea863b03cdb0302ac517728bc339456be91f23263bd8a41df73eaff1aa4","schema_version":"1.0","event_id":"sha256:55547ea863b03cdb0302ac517728bc339456be91f23263bd8a41df73eaff1aa4"}],"timestamp_proofs":[],"mirror_hints":[{"mirror_type":"https","name":"Pith Resolver","base_url":"https://pith.science","bundle_url":"https://pith.science/pith/6SRJ2EZQPXK4MVPFGEOE5YN3KE/bundle.json","state_url":"https://pith.science/pith/6SRJ2EZQPXK4MVPFGEOE5YN3KE/state.json","well_known_bundle_url":"https://pith.science/.well-known/pith/6SRJ2EZQPXK4MVPFGEOE5YN3KE/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":"2026-07-07T13:48:37Z","links":{"resolver":"https://pith.science/pith/6SRJ2EZQPXK4MVPFGEOE5YN3KE","bundle":"https://pith.science/pith/6SRJ2EZQPXK4MVPFGEOE5YN3KE/bundle.json","state":"https://pith.science/pith/6SRJ2EZQPXK4MVPFGEOE5YN3KE/state.json","well_known_bundle":"https://pith.science/.well-known/pith/6SRJ2EZQPXK4MVPFGEOE5YN3KE/bundle.json"},"state":{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2025:6SRJ2EZQPXK4MVPFGEOE5YN3KE","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":"6010c76888d9fae9c50a284ade283c029ae3454cf3e10c7bbee74b44ef551dff","cross_cats_sorted":["cs.CL"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-01-31T22:39:04Z","title_canon_sha256":"2689c97482820d5c2bd22a10a9951448b9bdb819dc5026599e021ed9023cceb0"},"schema_version":"1.0","source":{"id":"2502.00203","kind":"arxiv","version":2}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2502.00203","created_at":"2026-07-05T10:11:32Z"},{"alias_kind":"arxiv_version","alias_value":"2502.00203v2","created_at":"2026-07-05T10:11:32Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2502.00203","created_at":"2026-07-05T10:11:32Z"},{"alias_kind":"pith_short_12","alias_value":"6SRJ2EZQPXK4","created_at":"2026-07-05T10:11:32Z"},{"alias_kind":"pith_short_16","alias_value":"6SRJ2EZQPXK4MVPF","created_at":"2026-07-05T10:11:32Z"},{"alias_kind":"pith_short_8","alias_value":"6SRJ2EZQ","created_at":"2026-07-05T10:11:32Z"}],"graph_snapshots":[{"event_id":"sha256:55547ea863b03cdb0302ac517728bc339456be91f23263bd8a41df73eaff1aa4","target":"graph","created_at":"2026-07-05T10:11:32Z","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/2502.00203/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"The rapid development of large language model (LLM) alignment algorithms has resulted in a complex and fragmented landscape, with limited clarity on the effectiveness of different methods and their inter-connections. This paper introduces Reward-Aware Preference Optimization (RPO), a mathematical framework that unifies popular preference optimization techniques in LLM alignment, including DPO, IPO, SimPO, and REINFORCE (LOO), among others. RPO provides a structured approach to disentangle and systematically study the impact of various design choices, such as the optimization objective, the num","authors_text":"Adithya Renduchintala, Alexander Bukharin, David Mosallanezhad, Dmitry Chichkov, Gerald Shen, Jiaqi Zeng, Oleksii Kuchaiev, Olivier Delalleau, Shengyang Sun, Soumye Singhal, Tugrul Konuk, Yian Zhang, Yi Dong, Zhilin Wang","cross_cats":["cs.CL"],"headline":"","license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-01-31T22:39:04Z","title":"Reward-aware Preference Optimization: A Unified Mathematical Framework for Model Alignment"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2502.00203","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":"sha256:00ac48166d937f6e423e5e97947e8762ae3d61818d283015fe7230e9d2d23f04","target":"record","created_at":"2026-07-05T10:11:32Z","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":"6010c76888d9fae9c50a284ade283c029ae3454cf3e10c7bbee74b44ef551dff","cross_cats_sorted":["cs.CL"],"license":"http://creativecommons.org/licenses/by/4.0/","primary_cat":"cs.LG","submitted_at":"2025-01-31T22:39:04Z","title_canon_sha256":"2689c97482820d5c2bd22a10a9951448b9bdb819dc5026599e021ed9023cceb0"},"schema_version":"1.0","source":{"id":"2502.00203","kind":"arxiv","version":2}},"canonical_sha256":"f4a29d13307dd5c655e5311c4ee1bb51127e338dcdbd8353a0ae7274305984a1","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"f4a29d13307dd5c655e5311c4ee1bb51127e338dcdbd8353a0ae7274305984a1","first_computed_at":"2026-07-05T10:11:32.686495Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T10:11:32.686495Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"Qz8btzo9q2+zHpKRJw617f4kTT5e5DeFzmZH1dtWeZnzOKfR0bVYOXXUM6P6KxLmZeq+Xa51iqhQDn4S67f0CQ==","signature_status":"signed_v1","signed_at":"2026-07-05T10:11:32.687141Z","signed_message":"canonical_sha256_bytes"},"source_id":"2502.00203","source_kind":"arxiv","source_version":2}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:00ac48166d937f6e423e5e97947e8762ae3d61818d283015fe7230e9d2d23f04","sha256:55547ea863b03cdb0302ac517728bc339456be91f23263bd8a41df73eaff1aa4"],"state_sha256":"e0900f748285d999194e325b5ba944d45f9c11a9017eef12c2788f760286766e"},"bundle_signature":{"signature_status":"signed_v1","algorithm":"ed25519","key_id":"pith-v1-2026-05","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","signature_b64":"GQ9S6jQutKaLdF5scalJXU7cNs08ffVtfr6c4Pz1HeRgJ80sAXv8iVLWOZnIp5yR+IMv2S0mR2JQjMHxsJWHAg==","signed_message":"bundle_sha256_bytes","signed_at":"2026-07-07T13:48:37.596085Z","bundle_sha256":"d598eb91f267396c5e95653b3675c3d715c23e56c3a26cf3ffdcf9b96925d513"}}