{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2024:7LPQM4CBKI6IJLFXS4SPBULVWL","short_pith_number":"pith:7LPQM4CB","schema_version":"1.0","canonical_sha256":"fadf067041523c84acb79724f0d175b2cf4b4855d6bbde6f3cae47a8b2a89787","source":{"kind":"arxiv","id":"2405.07863","version":3},"attestation_state":"computed","paper":{"title":"RLHF Workflow: From Reward Modeling to Online RLHF","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"Online iterative RLHF using proxy preference models from open-source datasets reaches state-of-the-art results on LLM chatbot benchmarks.","cross_cats":["cs.AI","cs.CL","stat.ML"],"primary_cat":"cs.LG","authors_text":"Bo Pang, Caiming Xiong, Doyen Sahoo, HanZe Dong, Han Zhao, Haoxiang Wang, Nan Jiang, Tong Zhang, Wei Xiong, Yingbo Zhou","submitted_at":"2024-05-13T15:50:39Z","abstract_excerpt":"We present the workflow of Online Iterative Reinforcement Learning from Human Feedback (RLHF) in this technical report, which is widely reported to outperform its offline counterpart by a large margin in the recent large language model (LLM) literature. However, existing open-source RLHF projects are still largely confined to the offline learning setting. In this technical report, we aim to fill in this gap and provide a detailed recipe that is easy to reproduce for online iterative RLHF. In particular, since online human feedback is usually infeasible for open-source communities with limited "},"verification_status":{"content_addressed":true,"pith_receipt":true,"author_attested":false,"weak_author_claims":0,"strong_author_claims":0,"externally_anchored":false,"storage_verified":false,"citation_signatures":0,"replication_records":0,"graph_snapshot":true,"references_resolved":true,"formal_links_present":true},"canonical_record":{"source":{"id":"2405.07863","kind":"arxiv","version":3},"metadata":{"license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","primary_cat":"cs.LG","submitted_at":"2024-05-13T15:50:39Z","cross_cats_sorted":["cs.AI","cs.CL","stat.ML"],"title_canon_sha256":"9f4678f3878ae83510a56d06bd2ee031ccdaa6a9660221f34c78267ca1f6e80d","abstract_canon_sha256":"592586c7b486cb0c7461dffbf62ec428773c3a092eac9be768e4c02bd653d800"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-05-17T23:38:13.576666Z","signature_b64":"YcKqCIMAlZzWs3QUjs9PomQ9Vt25WerhXj2TU79FE7tPhNK9Gzwb+FfkwqK+Ykbx+NRg9I7x3xn0yLTBAh5PDA==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"fadf067041523c84acb79724f0d175b2cf4b4855d6bbde6f3cae47a8b2a89787","last_reissued_at":"2026-05-17T23:38:13.576059Z","signature_status":"signed_v1","first_computed_at":"2026-05-17T23:38:13.576059Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"RLHF Workflow: From Reward Modeling to Online RLHF","license":"http://arxiv.org/licenses/nonexclusive-distrib/1.0/","headline":"Online iterative RLHF using proxy preference models from open-source datasets reaches state-of-the-art results on LLM chatbot benchmarks.","cross_cats":["cs.AI","cs.CL","stat.ML"],"primary_cat":"cs.LG","authors_text":"Bo Pang, Caiming Xiong, Doyen Sahoo, HanZe Dong, Han Zhao, Haoxiang Wang, Nan Jiang, Tong Zhang, Wei Xiong, Yingbo Zhou","submitted_at":"2024-05-13T15:50:39Z","abstract_excerpt":"We present the workflow of Online Iterative Reinforcement Learning from Human Feedback (RLHF) in this technical report, which is widely reported to outperform its offline counterpart by a large margin in the recent large language model (LLM) literature. However, existing open-source RLHF projects are still largely confined to the offline learning setting. In this technical report, we aim to fill in this gap and provide a detailed recipe that is easy to reproduce for online iterative RLHF. In particular, since online human feedback is usually infeasible for open-source communities with limited "},"claims":{"count":4,"items":[{"kind":"strongest_claim","text":"We have shown that supervised fine-tuning (SFT) and iterative RLHF can obtain state-of-the-art performance with fully open-source datasets.","source":"verdict.strongest_claim","status":"machine_extracted","claim_id":"C1","attestation":"unclaimed"},{"kind":"weakest_assumption","text":"The proxy preference model built from open-source datasets approximates real human feedback closely enough that online RLHF updates remain beneficial rather than harmful.","source":"verdict.weakest_assumption","status":"machine_extracted","claim_id":"C2","attestation":"unclaimed"},{"kind":"one_line_summary","text":"The paper supplies a complete open-source recipe for online iterative RLHF that uses proxy preference models and reaches competitive performance on AlpacaEval-2, Arena-Hard, and MT-Bench.","source":"verdict.one_line_summary","status":"machine_extracted","claim_id":"C3","attestation":"unclaimed"},{"kind":"headline","text":"Online iterative RLHF using proxy preference models from open-source datasets reaches state-of-the-art results on LLM chatbot benchmarks.","source":"verdict.pith_extraction.headline","status":"machine_extracted","claim_id":"C4","attestation":"unclaimed"}],"snapshot_sha256":"e6fa12b1e95f7cd16c20a0d0019b40f432fbbd80191bf69f844eaaae261c014a"},"source":{"id":"2405.07863","kind":"arxiv","version":3},"verdict":{"id":"8c2a5daa-ca4f-4c5d-bf1d-b112b840e968","model_set":{"reader":"grok-4.3"},"created_at":"2026-05-17T16:43:20.020674Z","strongest_claim":"We have shown that supervised fine-tuning (SFT) and iterative RLHF can obtain state-of-the-art performance with fully open-source datasets.","one_line_summary":"The paper supplies a complete open-source recipe for online iterative RLHF that uses proxy preference models and reaches competitive performance on AlpacaEval-2, Arena-Hard, and MT-Bench.","pipeline_version":"pith-pipeline@v0.9.0","weakest_assumption":"The proxy preference model built from open-source datasets approximates real human feedback closely enough that online RLHF updates remain beneficial rather than harmful.","pith_extraction_headline":"Online iterative RLHF using proxy preference models from open-source datasets reaches state-of-the-art results on LLM chatbot benchmarks."},"references":{"count":23,"sample":[{"doi":"","year":null,"title":"Flat-leaf parsley is often preferred for its robust flavor","work_id":"3c2694e2-896c-47f0-8e99-d90dd293a9d5","ref_index":1,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":null,"title":"Garlic: Finely chopped or minced raw garlic gives a sharp, pungent taste to the Gremolata","work_id":"795ecb39-7a9d-45e9-87fe-a657ed7c6404","ref_index":2,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":null,"title":"In some variations, you might also find:","work_id":"34e57393-c913-4f41-beef-c1879ce1f78d","ref_index":3,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":2024,"title":"Lemon juice: This adds a little extra acidity and freshness to the mix. 5. Fresh basil or mint: These herbs can be used instead of, or in addition to, parsley, depending on the dish and personal prefe","work_id":"ce76a768-aba2-4aaa-b916-bcf98d0042c2","ref_index":4,"cited_arxiv_id":"","is_internal_anchor":false},{"doi":"","year":null,"title":"Squats: Helps to strengthen the lower body, particularly the glutes, quadriceps, and hamstrings. 2. Push-ups: Improves upper body strength, targeting the chest, triceps, and shoulders. 3. Planks: Enga","work_id":"f910d4d1-1e52-4117-91f2-beab3e5ef5b6","ref_index":5,"cited_arxiv_id":"","is_internal_anchor":false}],"resolved_work":23,"snapshot_sha256":"481a0d23db2a484444eb7d32b096bc5398924b4cdb6c5aa0656555238d98c04b","internal_anchors":0},"formal_canon":{"evidence_count":2,"snapshot_sha256":"d12b6b4e4011789984be2d919d3be1f95b12751953a42594f86cb7a4e637d156"},"author_claims":{"count":0,"strong_count":0,"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"builder_version":"pith-number-builder-2026-05-17-v1"},"aliases":[{"alias_kind":"arxiv","alias_value":"2405.07863","created_at":"2026-05-17T23:38:13.576164+00:00"},{"alias_kind":"arxiv_version","alias_value":"2405.07863v3","created_at":"2026-05-17T23:38:13.576164+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2405.07863","created_at":"2026-05-17T23:38:13.576164+00:00"},{"alias_kind":"pith_short_12","alias_value":"7LPQM4CBKI6I","created_at":"2026-05-18T12:33:37.589309+00:00"},{"alias_kind":"pith_short_16","alias_value":"7LPQM4CBKI6IJLFX","created_at":"2026-05-18T12:33:37.589309+00:00"},{"alias_kind":"pith_short_8","alias_value":"7LPQM4CB","created_at":"2026-05-18T12:33:37.589309+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":19,"internal_anchor_count":19,"sample":[{"citing_arxiv_id":"2509.23102","citing_title":"Multiplayer Nash Preference Optimization","ref_index":7,"is_internal_anchor":true},{"citing_arxiv_id":"2410.18451","citing_title":"Skywork-Reward: Bag of Tricks for Reward Modeling in LLMs","ref_index":8,"is_internal_anchor":true},{"citing_arxiv_id":"2412.14164","citing_title":"MetaMorph: Multimodal Understanding and Generation via Instruction Tuning","ref_index":164,"is_internal_anchor":true},{"citing_arxiv_id":"2406.16860","citing_title":"Cambrian-1: A Fully Open, Vision-Centric Exploration of Multimodal LLMs","ref_index":39,"is_internal_anchor":true},{"citing_arxiv_id":"2411.10442","citing_title":"Enhancing the Reasoning Ability of Multimodal Large Language Models via Mixed Preference Optimization","ref_index":26,"is_internal_anchor":true},{"citing_arxiv_id":"2602.01003","citing_title":"ESSAM: A Novel Competitive Evolution Strategies Approach to Reinforcement Learning for Memory Efficient LLMs Fine-Tuning","ref_index":2,"is_internal_anchor":true},{"citing_arxiv_id":"2602.12125","citing_title":"Learning beyond Teacher: Generalized On-Policy Distillation with Reward Extrapolation","ref_index":5,"is_internal_anchor":true},{"citing_arxiv_id":"2605.11679","citing_title":"Explaining and Breaking the Safety-Helpfulness Ceiling via Preference Dimensional Expansion","ref_index":50,"is_internal_anchor":true},{"citing_arxiv_id":"2604.01473","citing_title":"SelfGrader: Stable Jailbreak Detection for Large Language Models using Token-Level Logits","ref_index":8,"is_internal_anchor":true},{"citing_arxiv_id":"2605.11679","citing_title":"Explaining and Breaking the Safety-Helpfulness Ceiling via Preference Dimensional Expansion","ref_index":50,"is_internal_anchor":true},{"citing_arxiv_id":"2503.09567","citing_title":"Towards Reasoning Era: A Survey of Long Chain-of-Thought for Reasoning Large Language Models","ref_index":159,"is_internal_anchor":true},{"citing_arxiv_id":"2605.09119","citing_title":"Personalized Alignment Revisited: The Necessity and Sufficiency of User Diversity","ref_index":2,"is_internal_anchor":true},{"citing_arxiv_id":"2605.06036","citing_title":"Optimal Transport for LLM Reward Modeling from Noisy Preference","ref_index":259,"is_internal_anchor":true},{"citing_arxiv_id":"2605.01123","citing_title":"PERSA: Reinforcement Learning for Professor-Style Personalized Feedback with LLMs","ref_index":46,"is_internal_anchor":true},{"citing_arxiv_id":"2604.13035","citing_title":"SceneCritic: A Symbolic Evaluator for 3D Indoor Scene Synthesis","ref_index":13,"is_internal_anchor":true},{"citing_arxiv_id":"2605.07331","citing_title":"Rethinking Importance Sampling in LLM Policy Optimization: A Cumulative Token Perspective","ref_index":20,"is_internal_anchor":true},{"citing_arxiv_id":"2604.05341","citing_title":"Curr-RLCER:Curriculum Reinforcement Learning For Coherence Explainable Recommendation","ref_index":4,"is_internal_anchor":true},{"citing_arxiv_id":"2604.13602","citing_title":"Reward Hacking in the Era of Large Models: Mechanisms, Emergent Misalignment, Challenges","ref_index":35,"is_internal_anchor":true},{"citing_arxiv_id":"2604.16918","citing_title":"Freshness-Aware Prioritized Experience Replay for LLM/VLM Reinforcement Learning","ref_index":3,"is_internal_anchor":true}]},"formal_canon":{"evidence_count":2,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/7LPQM4CBKI6IJLFXS4SPBULVWL","json":"https://pith.science/pith/7LPQM4CBKI6IJLFXS4SPBULVWL.json","graph_json":"https://pith.science/api/pith-number/7LPQM4CBKI6IJLFXS4SPBULVWL/graph.json","events_json":"https://pith.science/api/pith-number/7LPQM4CBKI6IJLFXS4SPBULVWL/events.json","paper":"https://pith.science/paper/7LPQM4CB"},"agent_actions":{"view_html":"https://pith.science/pith/7LPQM4CBKI6IJLFXS4SPBULVWL","download_json":"https://pith.science/pith/7LPQM4CBKI6IJLFXS4SPBULVWL.json","view_paper":"https://pith.science/paper/7LPQM4CB","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2405.07863&json=true","fetch_graph":"https://pith.science/api/pith-number/7LPQM4CBKI6IJLFXS4SPBULVWL/graph.json","fetch_events":"https://pith.science/api/pith-number/7LPQM4CBKI6IJLFXS4SPBULVWL/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/7LPQM4CBKI6IJLFXS4SPBULVWL/action/timestamp_anchor","attest_storage":"https://pith.science/pith/7LPQM4CBKI6IJLFXS4SPBULVWL/action/storage_attestation","attest_author":"https://pith.science/pith/7LPQM4CBKI6IJLFXS4SPBULVWL/action/author_attestation","sign_citation":"https://pith.science/pith/7LPQM4CBKI6IJLFXS4SPBULVWL/action/citation_signature","submit_replication":"https://pith.science/pith/7LPQM4CBKI6IJLFXS4SPBULVWL/action/replication_record"}},"created_at":"2026-05-17T23:38:13.576164+00:00","updated_at":"2026-05-17T23:38:13.576164+00:00"}