{"state_type":"pith_open_graph_state","state_version":"1.0","pith_number":"pith:2024:MG44DPSWKC6QW63CHRNA42TICS","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":"425d36b99bbc2583586da7c8ac9e76b9eb52078c9c50553eb71ef37507f8b7e0","cross_cats_sorted":["cs.AI","cs.LG"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-06-17T17:55:38Z","title_canon_sha256":"a19c83a05c24259876e835be918f2fe8c27f2da2dbe6cfa22eec141da697dc20"},"schema_version":"1.0","source":{"id":"2406.11817","kind":"arxiv","version":1}},"source_aliases":[{"alias_kind":"arxiv","alias_value":"2406.11817","created_at":"2026-07-05T08:32:58Z"},{"alias_kind":"arxiv_version","alias_value":"2406.11817v1","created_at":"2026-07-05T08:32:58Z"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2406.11817","created_at":"2026-07-05T08:32:58Z"},{"alias_kind":"pith_short_12","alias_value":"MG44DPSWKC6Q","created_at":"2026-07-05T08:32:58Z"},{"alias_kind":"pith_short_16","alias_value":"MG44DPSWKC6QW63C","created_at":"2026-07-05T08:32:58Z"},{"alias_kind":"pith_short_8","alias_value":"MG44DPSW","created_at":"2026-07-05T08:32:58Z"}],"graph_snapshots":[{"event_id":"sha256:e6562f8268a2871391f0b8392c55cc463bae3167e679a3bd9cde879d32f13a64","target":"graph","created_at":"2026-07-05T08:32:58Z","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/2406.11817/integrity.json","findings":[],"snapshot_sha256":"c28c3603d3b5d939e8dc4c7e95fa8dfce3d595e45f758748cecf8e644a296938","summary":{"advisory":0,"by_detector":{},"critical":0,"informational":0}},"paper":{"abstract_excerpt":"Direct Preference Optimization (DPO), a standard method for aligning language models with human preferences, is traditionally applied to offline preferences. Recent studies show that DPO benefits from iterative training with online preferences labeled by a trained reward model. In this work, we identify a pitfall of vanilla iterative DPO - improved response quality can lead to increased verbosity. To address this, we introduce iterative length-regularized DPO (iLR-DPO) to penalize response length. Our empirical results show that iLR-DPO can enhance a 7B model to perform on par with GPT-4 witho","authors_text":"Chao Yang, Han-Sen Zhong, Jiaheng Liu, Jie Liu, Wanli Ouyang, Xingyuan Bu, Zhanhui Zhou","cross_cats":["cs.AI","cs.LG"],"headline":"","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-06-17T17:55:38Z","title":"Iterative Length-Regularized Direct Preference Optimization: A Case Study on Improving 7B Language Models to GPT-4 Level"},"references":{"count":0,"internal_anchors":0,"resolved_work":0,"sample":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2406.11817","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":"sha256:7b6a63d1cc8286010cf4daff55f056da37b0224490bcd38f22ac05c7411d996c","target":"record","created_at":"2026-07-05T08:32:58Z","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":"425d36b99bbc2583586da7c8ac9e76b9eb52078c9c50553eb71ef37507f8b7e0","cross_cats_sorted":["cs.AI","cs.LG"],"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.CL","submitted_at":"2024-06-17T17:55:38Z","title_canon_sha256":"a19c83a05c24259876e835be918f2fe8c27f2da2dbe6cfa22eec141da697dc20"},"schema_version":"1.0","source":{"id":"2406.11817","kind":"arxiv","version":1}},"canonical_sha256":"61b9c1be5650bd0b7b623c5a0e6a6814b10bdea2383059b792b2991a22b74a04","receipt":{"algorithm":"ed25519","builder_version":"pith-number-builder-2026-05-17-v1","canonical_sha256":"61b9c1be5650bd0b7b623c5a0e6a6814b10bdea2383059b792b2991a22b74a04","first_computed_at":"2026-07-05T08:32:58.274556Z","key_id":"pith-v1-2026-05","kind":"pith_receipt","last_reissued_at":"2026-07-05T08:32:58.274556Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54","receipt_version":"0.3","signature_b64":"SoUmmabDlwPxVaxhZcV68Y0WI/JG76LrSZW8Umm9Cy60rLDhBTMVg6Z47kp4ALsHiQe5u/1t2EuX3x2o/p6LCw==","signature_status":"signed_v1","signed_at":"2026-07-05T08:32:58.274976Z","signed_message":"canonical_sha256_bytes"},"source_id":"2406.11817","source_kind":"arxiv","source_version":1}}},"equivocations":[],"invalid_events":[],"applied_event_ids":["sha256:7b6a63d1cc8286010cf4daff55f056da37b0224490bcd38f22ac05c7411d996c","sha256:e6562f8268a2871391f0b8392c55cc463bae3167e679a3bd9cde879d32f13a64"],"state_sha256":"7e55ef9db4d7ddd7eeb70891f0449359aa0d6e07eb369b67abe1ccb2cf3e52b8"}