{"record_type":"pith_number_record","schema_url":"https://pith.science/schemas/pith-number/v1.json","pith_number":"pith:2023:3B3QFTDF427RD6CWRQLMAGXBOF","short_pith_number":"pith:3B3QFTDF","schema_version":"1.0","canonical_sha256":"d87702cc65e6bf11f8568c16c01ae17144ee76c93ac559c12e0717da1fb88955","source":{"kind":"arxiv","id":"2310.10505","version":4},"attestation_state":"computed","paper":{"title":"ReMax: A Simple, Effective, and Efficient Reinforcement Learning Method for Aligning Large Language Models","license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Ruoyu Sun, Tian Xu, Yang Yu, Yushun Zhang, Zhihang Lin, Zhi-Quan Luo, Ziniu Li","submitted_at":"2023-10-16T15:25:14Z","abstract_excerpt":"Reinforcement Learning from Human Feedback (RLHF) is key to aligning Large Language Models (LLMs), typically paired with the Proximal Policy Optimization (PPO) algorithm. While PPO is a powerful method designed for general reinforcement learning tasks, it is overly sophisticated for LLMs, leading to laborious hyper-parameter tuning and significant computation burdens. To make RLHF efficient, we present ReMax, which leverages 3 properties of RLHF: fast simulation, deterministic transitions, and trajectory-level rewards. These properties are not exploited in PPO, making it less suitable for RLHF"},"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":false,"formal_links_present":false},"canonical_record":{"source":{"id":"2310.10505","kind":"arxiv","version":4},"metadata":{"license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","primary_cat":"cs.LG","submitted_at":"2023-10-16T15:25:14Z","cross_cats_sorted":[],"title_canon_sha256":"569d14f60508783cfd4a3bf94408309cc180689670d80c03deb421ce2692b633","abstract_canon_sha256":"99522b91743a4d84326fc86980148d6b5cdfee0dad1857aab5acf4a345321c8e"},"schema_version":"1.0"},"receipt":{"kind":"pith_receipt","key_id":"pith-v1-2026-05","algorithm":"ed25519","signed_at":"2026-07-05T08:19:37.655584Z","signature_b64":"F1u0GXzD/1dlnKw/uIMGL62c8IL2mhwu87i2Gby6GaAPkoLH16X7O01oPKpxRcFqW+7VjTn6gnJM0y3d2Zj/AQ==","signed_message":"canonical_sha256_bytes","builder_version":"pith-number-builder-2026-05-17-v1","receipt_version":"0.3","canonical_sha256":"d87702cc65e6bf11f8568c16c01ae17144ee76c93ac559c12e0717da1fb88955","last_reissued_at":"2026-07-05T08:19:37.655148Z","signature_status":"signed_v1","first_computed_at":"2026-07-05T08:19:37.655148Z","public_key_fingerprint":"8d4b5ee74e4693bcd1df2446408b0d54"},"graph_snapshot":{"paper":{"title":"ReMax: A Simple, Effective, and Efficient Reinforcement Learning Method for Aligning Large Language Models","license":"http://creativecommons.org/licenses/by-nc-nd/4.0/","headline":"","cross_cats":[],"primary_cat":"cs.LG","authors_text":"Ruoyu Sun, Tian Xu, Yang Yu, Yushun Zhang, Zhihang Lin, Zhi-Quan Luo, Ziniu Li","submitted_at":"2023-10-16T15:25:14Z","abstract_excerpt":"Reinforcement Learning from Human Feedback (RLHF) is key to aligning Large Language Models (LLMs), typically paired with the Proximal Policy Optimization (PPO) algorithm. While PPO is a powerful method designed for general reinforcement learning tasks, it is overly sophisticated for LLMs, leading to laborious hyper-parameter tuning and significant computation burdens. To make RLHF efficient, we present ReMax, which leverages 3 properties of RLHF: fast simulation, deterministic transitions, and trajectory-level rewards. These properties are not exploited in PPO, making it less suitable for RLHF"},"claims":{"count":0,"items":[],"snapshot_sha256":"258153158e38e3291e3d48162225fcdb2d5a3ed65a07baac614ab91432fd4f57"},"source":{"id":"2310.10505","kind":"arxiv","version":4},"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/2310.10505/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"},"aliases":[{"alias_kind":"arxiv","alias_value":"2310.10505","created_at":"2026-07-05T08:19:37.655213+00:00"},{"alias_kind":"arxiv_version","alias_value":"2310.10505v4","created_at":"2026-07-05T08:19:37.655213+00:00"},{"alias_kind":"doi","alias_value":"10.48550/arxiv.2310.10505","created_at":"2026-07-05T08:19:37.655213+00:00"},{"alias_kind":"pith_short_12","alias_value":"3B3QFTDF427R","created_at":"2026-07-05T08:19:37.655213+00:00"},{"alias_kind":"pith_short_16","alias_value":"3B3QFTDF427RD6CW","created_at":"2026-07-05T08:19:37.655213+00:00"},{"alias_kind":"pith_short_8","alias_value":"3B3QFTDF","created_at":"2026-07-05T08:19:37.655213+00:00"}],"events":[],"event_summary":{},"paper_claims":[],"inbound_citations":{"count":20,"internal_anchor_count":0,"sample":[{"citing_arxiv_id":"2606.21943","citing_title":"Modularized Reinforcement Learning on LLMs: From MDP Creation to Exploration and Learning","ref_index":111,"is_internal_anchor":false},{"citing_arxiv_id":"2606.17250","citing_title":"Rethinking Groups in Critic-Free RLVR","ref_index":26,"is_internal_anchor":false},{"citing_arxiv_id":"2606.01249","citing_title":"Trust Region On-Policy Distillation","ref_index":215,"is_internal_anchor":false},{"citing_arxiv_id":"2605.27736","citing_title":"Explicit Critic Guidance for Aligning Diffusion Models","ref_index":37,"is_internal_anchor":false},{"citing_arxiv_id":"2605.28388","citing_title":"Mechanistically Interpreting the Role of Sample Difficulty in RLVR for LLMs","ref_index":20,"is_internal_anchor":false},{"citing_arxiv_id":"2605.06139","citing_title":"Listwise Policy Optimization: Group-based RLVR as Target-Projection on the LLM Response Simplex","ref_index":46,"is_internal_anchor":false},{"citing_arxiv_id":"2605.20277","citing_title":"Regulating Anatomy-Aware Rewards via Trajectory-Integral Feedback for Volumetric Computed Tomography Analysis","ref_index":48,"is_internal_anchor":false},{"citing_arxiv_id":"2605.17497","citing_title":"Self-Supervised On-Policy Distillation for Reasoning Language Models","ref_index":99,"is_internal_anchor":false},{"citing_arxiv_id":"2605.19282","citing_title":"Rethinking Muon Beyond Pretraining: Spectral Failures and High-Pass Remedies for VLA and RLVR","ref_index":20,"is_internal_anchor":false},{"citing_arxiv_id":"2509.08827","citing_title":"A Survey of Reinforcement Learning for Large Reasoning Models","ref_index":297,"is_internal_anchor":false},{"citing_arxiv_id":"2512.13592","citing_title":"Image Diffusion Preview with Consistency Solver","ref_index":20,"is_internal_anchor":false},{"citing_arxiv_id":"2601.21972","citing_title":"Learning Decentralized LLM Collaboration with Multi-Agent Actor Critic","ref_index":18,"is_internal_anchor":false},{"citing_arxiv_id":"2605.11235","citing_title":"Internalizing Curriculum Judgment for LLM Reinforcement Fine-Tuning","ref_index":20,"is_internal_anchor":false},{"citing_arxiv_id":"2605.08666","citing_title":"The Cancellation Hypothesis in Critic-Free RL: From Outcome Rewards to Token Credits","ref_index":10,"is_internal_anchor":false},{"citing_arxiv_id":"2604.24273","citing_title":"BitRL: Reinforcement Learning with 1-bit Quantized Language Models for Resource-Constrained Edge Deployment","ref_index":20,"is_internal_anchor":false},{"citing_arxiv_id":"2502.01456","citing_title":"Process Reinforcement through Implicit Rewards","ref_index":97,"is_internal_anchor":false},{"citing_arxiv_id":"2605.06139","citing_title":"Listwise Policy Optimization: Group-based RLVR as Target-Projection on the LLM Response Simplex","ref_index":45,"is_internal_anchor":false},{"citing_arxiv_id":"2604.13010","citing_title":"Lightning OPD: Efficient Post-Training for Large Reasoning Models with Offline On-Policy Distillation","ref_index":24,"is_internal_anchor":false},{"citing_arxiv_id":"2409.19256","citing_title":"HybridFlow: A Flexible and Efficient RLHF Framework","ref_index":49,"is_internal_anchor":false},{"citing_arxiv_id":"2604.13010","citing_title":"Lightning OPD: Efficient Post-Training for Large Reasoning Models with Offline On-Policy Distillation","ref_index":24,"is_internal_anchor":false}]},"formal_canon":{"evidence_count":0,"sample":[],"anchors":[]},"links":{"html":"https://pith.science/pith/3B3QFTDF427RD6CWRQLMAGXBOF","json":"https://pith.science/pith/3B3QFTDF427RD6CWRQLMAGXBOF.json","graph_json":"https://pith.science/api/pith-number/3B3QFTDF427RD6CWRQLMAGXBOF/graph.json","events_json":"https://pith.science/api/pith-number/3B3QFTDF427RD6CWRQLMAGXBOF/events.json","paper":"https://pith.science/paper/3B3QFTDF"},"agent_actions":{"view_html":"https://pith.science/pith/3B3QFTDF427RD6CWRQLMAGXBOF","download_json":"https://pith.science/pith/3B3QFTDF427RD6CWRQLMAGXBOF.json","view_paper":"https://pith.science/paper/3B3QFTDF","resolve_alias":"https://pith.science/api/pith-number/resolve?arxiv=2310.10505&json=true","fetch_graph":"https://pith.science/api/pith-number/3B3QFTDF427RD6CWRQLMAGXBOF/graph.json","fetch_events":"https://pith.science/api/pith-number/3B3QFTDF427RD6CWRQLMAGXBOF/events.json","actions":{"anchor_timestamp":"https://pith.science/pith/3B3QFTDF427RD6CWRQLMAGXBOF/action/timestamp_anchor","attest_storage":"https://pith.science/pith/3B3QFTDF427RD6CWRQLMAGXBOF/action/storage_attestation","attest_author":"https://pith.science/pith/3B3QFTDF427RD6CWRQLMAGXBOF/action/author_attestation","sign_citation":"https://pith.science/pith/3B3QFTDF427RD6CWRQLMAGXBOF/action/citation_signature","submit_replication":"https://pith.science/pith/3B3QFTDF427RD6CWRQLMAGXBOF/action/replication_record"}},"created_at":"2026-07-05T08:19:37.655213+00:00","updated_at":"2026-07-05T08:19:37.655213+00:00"}