{"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/improving-policy-learning-via-language","title":"Improving Policy Learning via Language Dynamics Distillation","arxiv_id":"2210.00066","date":"2022-09-30","proceeding":null,"authors":["Victor Zhong","Jesse Mu","Luke Zettlemoyer","Edward Grefenstette","Tim Rocktäschel"],"abstract":"Recent work has shown that augmenting environments with language descriptions improves policy learning. However, for environments with complex language abstractions, learning how to ground language to observations is difficult due to sparse, delayed rewards. We propose Language Dynamics Distillation (LDD), which pretrains a model to predict environment dynamics given demonstrations with language descriptions, and then fine-tunes these language-aware pretrained representations via reinforcement learning (RL). In this way, the model is trained to both maximize expected reward and retain knowledge about how language relates to environment dynamics. On SILG, a benchmark of five tasks with language descriptions that evaluate distinct generalization challenges on unseen environments (NetHack, ALFWorld, RTFM, Messenger, and Touchdown), LDD outperforms tabula-rasa RL, VAE pretraining, and methods that learn from unlabeled demonstrations in inverse RL and reward shaping with pretrained experts. In our analyses, we show that language descriptions in demonstrations improve sample-efficiency and generalization across environments, and that dynamics modelling with expert demonstrations is more effective than with non-experts.","url_abs":"https://arxiv.org/abs/2210.00066v1","url_pdf":"https://arxiv.org/pdf/2210.00066v1.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":"improving-policy-learning-via-language","repo_url":"https://github.com/vzhong/language-dynamics-distillation","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":0,"framework":"pytorch","reach":{"status":"ok","spdx":"MIT"}}],"tasks":[{"task_slug":"nethack","task_name":"NetHack"},{"task_slug":"reinforcement-learning-1","task_name":"Reinforcement Learning (RL)"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=2210.00066","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2210.00066"}},"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":"deterministic:regex_extraction","url":"https://github.com/facebookresearch/nle","reach":null},{"provenance":"deterministic:regex_extraction","url":"https://github.com/vzhong/silg","reach":null},{"provenance":"deterministic:regex_extraction","url":"https://github.com/vzhong/language-dynamics-distillation","reach":{"status":"ok","spdx":"MIT"}},{"provenance":"deterministic:regex_extraction","url":"https://github.com/ahjwang/messenger-emma","reach":null}],"summary":{"ran":1,"ran_draft_wrong":1,"unverified":6},"by_repo_kind":{"official":{"samples":1,"ran":0,"repositories":1},"found_in_text":{"samples":5,"ran":2,"repositories":2}},"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":4,"samples":[{"code_sha256_prefix":"7d377f889cf16d0e","entry":"Encoder","repo":"ahjwang/messenger-emma","repo_kind":"found_in_text","path":"messenger/models/emma.py","file_url":"https://github.com/ahjwang/messenger-emma/blob/HEAD/messenger/models/emma.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"7d377f889cf16d0e"}},{"code_sha256_prefix":"dacaf8ab3063a9ff","entry":"nonzero_mean","repo":"ahjwang/messenger-emma","repo_kind":"found_in_text","path":"messenger/models/emma.py","file_url":"https://github.com/ahjwang/messenger-emma/blob/HEAD/messenger/models/emma.py","link_basis":"first_harvest_node","language":"python","status":"ran_draft_wrong","verification_level":1,"contract_check":"MISDECLARED","metamorphic_tier":null,"behaviour_fingerprint":true,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"dacaf8ab3063a9ff"}},{"code_sha256_prefix":"801be0895c4e0bf2","entry":"EMMA","repo":"ahjwang/messenger-emma","repo_kind":"found_in_text","path":"messenger/models/emma.py","file_url":"https://github.com/ahjwang/messenger-emma/blob/HEAD/messenger/models/emma.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"801be0895c4e0bf2"}},{"code_sha256_prefix":"e624e7bc82f85ecd","entry":"compute_baseline_loss","repo":null,"repo_kind":null,"path":null,"file_url":null,"link_basis":"identical_code_first_harvested_elsewhere","language":"python","status":"unverified","verification_level":0,"contract_check":"RAISES","metamorphic_tier":null,"behaviour_fingerprint":false,"licence":null,"inline_ok":false,"mcp_get_code":{"code_sha256":"e624e7bc82f85ecd"}},{"code_sha256_prefix":"3fba94f2f8bc0565","entry":"compute_entropy_loss","repo":null,"repo_kind":null,"path":null,"file_url":null,"link_basis":"identical_code_first_harvested_elsewhere","language":"python","status":"unverified","verification_level":0,"contract_check":"RAISES","metamorphic_tier":null,"behaviour_fingerprint":false,"licence":null,"inline_ok":false,"mcp_get_code":{"code_sha256":"3fba94f2f8bc0565"}},{"code_sha256_prefix":"f46896c3a13fa416","entry":"compute_policy_gradient_loss","repo":"facebookresearch/nle","repo_kind":"found_in_text","path":"nle/agent/agent.py","file_url":"https://github.com/facebookresearch/nle/blob/HEAD/nle/agent/agent.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NOASSERTION","inline_ok":false,"mcp_get_code":{"code_sha256":"f46896c3a13fa416"}},{"code_sha256_prefix":"fa8a88fb35b3b6f6","entry":"learn","repo":"facebookresearch/nle","repo_kind":"found_in_text","path":"nle/agent/agent.py","file_url":"https://github.com/facebookresearch/nle/blob/HEAD/nle/agent/agent.py","link_basis":"first_harvest_node","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NOASSERTION","inline_ok":false,"mcp_get_code":{"code_sha256":"fa8a88fb35b3b6f6"}},{"code_sha256_prefix":"22e55c2b5b19dab7","entry":"nonzero_mean","repo":"vzhong/language-dynamics-distillation","repo_kind":"official","path":"messenger/emma.py","file_url":"https://github.com/vzhong/language-dynamics-distillation/blob/HEAD/messenger/emma.py","link_basis":"harvester_set","language":"python","status":"unverified","verification_level":0,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"MIT","inline_ok":true,"mcp_get_code":{"code_sha256":"22e55c2b5b19dab7"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}