{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/entropy-regularized-token-level-policy","title":"Entropy-Regularized Token-Level Policy Optimization for Language Agent Reinforcement","arxiv_id":"2402.06700","date":"2024-02-09","proceeding":null,"authors":["Muning Wen","Junwei Liao","Cheng Deng","Jun Wang","Weinan Zhang","Ying Wen"],"abstract":"Large Language Models (LLMs) have shown promise as intelligent agents in interactive decision-making tasks. Traditional approaches often depend on meticulously designed prompts, high-quality examples, or additional reward models for in-context learning, supervised fine-tuning, or RLHF. Reinforcement learning (RL) presents a dynamic alternative for LLMs to overcome these dependencies by engaging directly with task-specific environments. Nonetheless, it faces significant hurdles: 1) instability stemming from the exponentially vast action space requiring exploration; 2) challenges in assigning token-level credit based on action-level reward signals, resulting in discord between maximizing rewards and accurately modeling corpus data. In response to these challenges, we introduce Entropy-Regularized Token-level Policy Optimization (ETPO), an entropy-augmented RL method tailored for optimizing LLMs at the token level. At the heart of ETPO is our novel per-token soft Bellman update, designed to harmonize the RL process with the principles of language modeling. This methodology decomposes the Q-function update from a coarse action-level view to a more granular token-level perspective, backed by theoretical proof of optimization consistency. Crucially, this decomposition renders linear time complexity in action exploration. We assess the effectiveness of ETPO within a simulated environment that models data science code generation as a series of multi-step interactive tasks; results underline ETPO's potential as a robust method for refining the interactive decision-making capabilities of language agents. For a more detailed preliminary work describing our motivation for token-level decomposition and applying it in PPO methods, please refer to arXiv:2405.15821.","url_abs":"https://arxiv.org/abs/2402.06700v4","url_pdf":"https://arxiv.org/pdf/2402.06700v4.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"entropy-regularized-token-level-policy","repo_url":"https://github.com/morning9393/etpo","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok"}}],"tasks":[{"task_slug":"code-generation","task_name":"Code Generation"},{"task_slug":"decision-making","task_name":"Decision Making"},{"task_slug":"in-context-learning","task_name":"In-Context Learning"},{"task_slug":"language-modeling","task_name":"Language Modeling"},{"task_slug":"language-modelling","task_name":"Language Modelling"},{"task_slug":"reinforcement-learning-1","task_name":"Reinforcement Learning (RL)"}],"methods":[{"method_slug":"entropy-regularization","method_name":"Entropy Regularization"},{"method_slug":"ppo","method_name":"PPO"}],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=2402.06700","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2402.06700"}},"developers":"https://syntology.ai/developers","read_at":"2026-09-24T18:15:14+00:00","read_at_is":"when the build read Syntology's graph, not when any sample ran","claim":"Per-sample execution status on synthesized fixtures; not a correctness claim about the paper. Samples come from repositories linked to the paper, official or community; repo_kind says which.","repos":[{"provenance":"external:paperswithcode_snapshot_2025-07-28","url":"https://github.com/morning9393/etpo","reach":{"status":"ok"}}],"summary":{"ran":2,"ran_draft_wrong":2,"ran_violates":1,"ran_fixture":1,"unverified":3},"by_repo_kind":{"official":{"samples":9,"ran":6,"repositories":1}},"repo_kind_vocabulary":{"official":"The archive marks this repository official for the paper","named_in_paper":"The archive records that the paper mentions this repository; it is not marked official","listed":"In the archive's code links for this paper, not marked official and not recorded as mentioned in the paper","found_in_text":"Syntology found this repository in the paper's own text; whether it is the authors' implementation is not asserted","community":"Not in the archive's code links for this paper; a community repository Syntology harvested"},"n_pointer_only_for_licence":9,"samples":[{"code_sha256_prefix":"1136e568a8e6dd2e","entry":"check","repo":"morning9393/etpo","repo_kind":"official","path":"etpo/utils/util.py","file_url":"https://github.com/morning9393/etpo/blob/HEAD/etpo/utils/util.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"1136e568a8e6dd2e"}},{"code_sha256_prefix":"8da73b3ead760a3d","entry":"get_gard_norm","repo":"morning9393/etpo","repo_kind":"official","path":"etpo/utils/util.py","file_url":"https://github.com/morning9393/etpo/blob/HEAD/etpo/utils/util.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"8da73b3ead760a3d"}},{"code_sha256_prefix":"317a7486c52962a6","entry":"huber_loss","repo":"morning9393/etpo","repo_kind":"official","path":"etpo/utils/util.py","file_url":"https://github.com/morning9393/etpo/blob/HEAD/etpo/utils/util.py","link_basis":"plan_row","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"MISDECLARED","metamorphic_tier":null,"behaviour_fingerprint":true,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"317a7486c52962a6"}},{"code_sha256_prefix":"e93cc5b705c2eb3b","entry":"precompute_freqs_cis","repo":"morning9393/etpo","repo_kind":"official","path":"etpo/models/codellama/model.py","file_url":"https://github.com/morning9393/etpo/blob/HEAD/etpo/models/codellama/model.py","link_basis":"harvester_set","language":"python","status":"ran_violates","verification_level":1,"contract_check":"VIOLATES","metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"e93cc5b705c2eb3b"}},{"code_sha256_prefix":"70bf6ebaafd266c4","entry":"reshape_for_broadcast","repo":"morning9393/etpo","repo_kind":"official","path":"etpo/models/codellama/model.py","file_url":"https://github.com/morning9393/etpo/blob/HEAD/etpo/models/codellama/model.py","link_basis":"harvester_set","language":"python","status":"ran_fixture","verification_level":1,"contract_check":"RAISES","metamorphic_tier":"invariant","behaviour_fingerprint":true,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"70bf6ebaafd266c4"}},{"code_sha256_prefix":"8845976729f4c4ee","entry":"sample_top_p","repo":"morning9393/etpo","repo_kind":"official","path":"etpo/models/codellama/generation.py","file_url":"https://github.com/morning9393/etpo/blob/HEAD/etpo/models/codellama/generation.py","link_basis":"harvester_set","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"MISDECLARED","metamorphic_tier":null,"behaviour_fingerprint":true,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"8845976729f4c4ee"}},{"code_sha256_prefix":"ed8b68645cfd07e0","entry":"apply_rotary_emb","repo":"morning9393/etpo","repo_kind":"official","path":"etpo/models/codellama/model.py","file_url":"https://github.com/morning9393/etpo/blob/HEAD/etpo/models/codellama/model.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"ed8b68645cfd07e0"}},{"code_sha256_prefix":"e7a5d8632aa5c7dd","entry":"load_dataset","repo":"morning9393/etpo","repo_kind":"official","path":"etpo/envs/datascience/scikit_env.py","file_url":"https://github.com/morning9393/etpo/blob/HEAD/etpo/envs/datascience/scikit_env.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"e7a5d8632aa5c7dd"}},{"code_sha256_prefix":"a67fe2b3be0765b7","entry":"load_kidney_stone","repo":"morning9393/etpo","repo_kind":"official","path":"etpo/envs/datascience/data_utils.py","file_url":"https://github.com/morning9393/etpo/blob/HEAD/etpo/envs/datascience/data_utils.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"a67fe2b3be0765b7"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}