{"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/the-generalization-gap-in-offline","title":"The Generalization Gap in Offline Reinforcement Learning","arxiv_id":"2312.05742","date":"2023-12-10","proceeding":null,"authors":["Ishita Mediratta","Qingfei You","Minqi Jiang","Roberta Raileanu"],"abstract":"Despite recent progress in offline learning, these methods are still trained and tested on the same environment. In this paper, we compare the generalization abilities of widely used online and offline learning methods such as online reinforcement learning (RL), offline RL, sequence modeling, and behavioral cloning. Our experiments show that offline learning algorithms perform worse on new environments than online learning ones. We also introduce the first benchmark for evaluating generalization in offline learning, collecting datasets of varying sizes and skill-levels from Procgen (2D video games) and WebShop (e-commerce websites). The datasets contain trajectories for a limited number of game levels or natural language instructions and at test time, the agent has to generalize to new levels or instructions. Our experiments reveal that existing offline learning algorithms struggle to match the performance of online RL on both train and test environments. Behavioral cloning is a strong baseline, outperforming state-of-the-art offline RL and sequence modeling approaches when trained on data from multiple environments and tested on new ones. Finally, we find that increasing the diversity of the data, rather than its size, improves performance on new environments for all offline learning algorithms. Our study demonstrates the limited generalization of current offline learning algorithms highlighting the need for more research in this area.","url_abs":"https://arxiv.org/abs/2312.05742v2","url_pdf":"https://arxiv.org/pdf/2312.05742v2.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":"the-generalization-gap-in-offline","repo_url":"https://github.com/facebookresearch/gen_dgrl","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"NOASSERTION"}}],"tasks":[{"task_slug":"offline-rl","task_name":"Offline RL"},{"task_slug":"reinforcement-learning","task_name":"Reinforcement Learning"},{"task_slug":"reinforcement-learning-1","task_name":"Reinforcement Learning (RL)"},{"task_slug":"reinforcement-learning-2","task_name":"reinforcement-learning"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=2312.05742","mcp":{"get_harvested_code_for_paper":{"arxiv_id":"2312.05742"}},"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/facebookresearch/gen_dgrl","reach":{"status":"ok","spdx":"NOASSERTION"}}],"summary":{"ran":7,"ran_fixture":1,"unverified":3},"by_repo_kind":{"official":{"samples":11,"ran":8,"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":11,"samples":[{"code_sha256_prefix":"edbb710d8687df70","entry":"arg2str","repo":"facebookresearch/gen_dgrl","repo_kind":"official","path":"webShop/baseline_models/slurm.py","file_url":"https://github.com/facebookresearch/gen_dgrl/blob/HEAD/webShop/baseline_models/slurm.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NOASSERTION","inline_ok":false,"mcp_get_code":{"code_sha256":"edbb710d8687df70"}},{"code_sha256_prefix":"658ecf571b8a449d","entry":"build_arg_parser","repo":"facebookresearch/gen_dgrl","repo_kind":"official","path":"procgen/download.py","file_url":"https://github.com/facebookresearch/gen_dgrl/blob/HEAD/procgen/download.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NOASSERTION","inline_ok":false,"mcp_get_code":{"code_sha256":"658ecf571b8a449d"}},{"code_sha256_prefix":"6c873c43e9aed0a7","entry":"discount_reward","repo":"facebookresearch/gen_dgrl","repo_kind":"official","path":"webShop/baseline_models/agent.py","file_url":"https://github.com/facebookresearch/gen_dgrl/blob/HEAD/webShop/baseline_models/agent.py","link_basis":"plan_row","language":"python","status":"ran_fixture","verification_level":1,"contract_check":"RAISES","metamorphic_tier":"invariant","behaviour_fingerprint":false,"licence":"NONE","inline_ok":false,"mcp_get_code":{"code_sha256":"6c873c43e9aed0a7"}},{"code_sha256_prefix":"a081306c38dcb0d3","entry":"new_process_goal","repo":"facebookresearch/gen_dgrl","repo_kind":"official","path":"webShop/baseline_models/human_trajectories_collector.py","file_url":"https://github.com/facebookresearch/gen_dgrl/blob/HEAD/webShop/baseline_models/human_trajectories_collector.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":true,"licence":"NOASSERTION","inline_ok":false,"mcp_get_code":{"code_sha256":"a081306c38dcb0d3"}},{"code_sha256_prefix":"cc483cf670852845","entry":"process","repo":"facebookresearch/gen_dgrl","repo_kind":"official","path":"webShop/baseline_models/il_data_collector.py","file_url":"https://github.com/facebookresearch/gen_dgrl/blob/HEAD/webShop/baseline_models/il_data_collector.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":true,"licence":"NOASSERTION","inline_ok":false,"mcp_get_code":{"code_sha256":"cc483cf670852845"}},{"code_sha256_prefix":"8012f9ad06833123","entry":"process_goal","repo":"facebookresearch/gen_dgrl","repo_kind":"official","path":"webShop/baseline_models/il_data_collector.py","file_url":"https://github.com/facebookresearch/gen_dgrl/blob/HEAD/webShop/baseline_models/il_data_collector.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":true,"licence":"NOASSERTION","inline_ok":false,"mcp_get_code":{"code_sha256":"8012f9ad06833123"}},{"code_sha256_prefix":"98948fae5be7e92e","entry":"profile","repo":"facebookresearch/gen_dgrl","repo_kind":"official","path":"webShop/baseline_models/logger.py","file_url":"https://github.com/facebookresearch/gen_dgrl/blob/HEAD/webShop/baseline_models/logger.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":false,"licence":"NOASSERTION","inline_ok":false,"mcp_get_code":{"code_sha256":"98948fae5be7e92e"}},{"code_sha256_prefix":"30817084576ab0fd","entry":"strip_website","repo":"facebookresearch/gen_dgrl","repo_kind":"official","path":"webShop/baseline_models/human_trajectories_collector.py","file_url":"https://github.com/facebookresearch/gen_dgrl/blob/HEAD/webShop/baseline_models/human_trajectories_collector.py","link_basis":"first_harvest_node","language":"python","status":"ran","verification_level":1,"contract_check":null,"metamorphic_tier":null,"behaviour_fingerprint":true,"licence":"NOASSERTION","inline_ok":false,"mcp_get_code":{"code_sha256":"30817084576ab0fd"}},{"code_sha256_prefix":"6ac021fec5c8d2c2","entry":"get_data","repo":"facebookresearch/gen_dgrl","repo_kind":"official","path":"webShop/baseline_models/train_choice_bcq.py","file_url":"https://github.com/facebookresearch/gen_dgrl/blob/HEAD/webShop/baseline_models/train_choice_bcq.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":"6ac021fec5c8d2c2"}},{"code_sha256_prefix":"e69c2e4d0254e7de","entry":"make_output_format","repo":"facebookresearch/gen_dgrl","repo_kind":"official","path":"webShop/baseline_models/logger.py","file_url":"https://github.com/facebookresearch/gen_dgrl/blob/HEAD/webShop/baseline_models/logger.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":"e69c2e4d0254e7de"}},{"code_sha256_prefix":"88ef5a17fc986d9a","entry":"read_json","repo":"facebookresearch/gen_dgrl","repo_kind":"official","path":"webShop/baseline_models/logger.py","file_url":"https://github.com/facebookresearch/gen_dgrl/blob/HEAD/webShop/baseline_models/logger.py","link_basis":"plan_row","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":"88ef5a17fc986d9a"}}]},"arxiv_metadata":null,"syntology_extracted_results":null}