{"url":"/method/go-explore","slug":"go-explore","name":"Go-Explore","full_name":"Go-Explore","full_name_withheld":false,"description_markdown":"**Go-Explore** is a family of algorithms aiming to tackle two challenges with effective exploration in reinforcement learning: algorithms forgetting how to reach previously visited states (\"detachment\") and from failing to first return to a state before exploring from it (\"derailment\").\r\n\r\nTo avoid detachment, Go-Explore builds an archive of the different states it has visited in the environment, thus ensuring that states cannot be forgotten. Starting with an archive beginning with the initial state, the archive is built iteratively. In Go-Explore we:\r\n\r\n(a) Probabilistically select a state from the archive, preferring states associated with promising cells. \r\n\r\n(b) Return to the selected state, such as by restoring simulator state or by running a goal-conditioned policy. \r\n\r\n(c) Explore from that state by taking random actions or sampling from a trained policy. \r\n\r\n(d) Map every state encountered during returning and exploring to a low-dimensional cell representation. \r\n\r\n(e) Add states that map to new cells to the archive and update other archive entries.","description_state":"present","introduced_year":null,"introduced_by":{"title":"Go-Explore: a New Approach for Hard-Exploration Problems","paper":"/paper/go-explore-a-new-approach-for-hard","first_author":"Adrien Ecoffet","n_authors":5,"url_abs":null,"archive_paper_url":"https://paperswithcode.com/paper/go-explore-a-new-approach-for-hard"},"source":{"url":"https://arxiv.org/abs/1901.10995v4","title":"Go-Explore: a New Approach for Hard-Exploration Problems","url_on_a_paper_host":true},"code_snippet_url":null,"code_snippet_url_on_a_code_host":false,"categories":[{"area":"Reinforcement Learning","area_id":"reinforcement-learning","collection":"Behaviour Policies","url":"/methods/category/behaviour-policies","pwc_aliases":[]}],"n_papers_tagged":16,"archive_num_papers":16,"papers_newest_first":[{"paper":"/paper/intelligent-go-explore-standing-on-the","title":"Intelligent Go-Explore: Standing on the Shoulders of Giant Foundation Models","date":"2024-05-24","arxiv_id":"2405.15143","n_code_links":1,"syntology":{"ran":2,"of":2,"unverified":0,"pointer_only":0}},{"paper":null,"title":"Go-Explore for Residential Energy Management","date":"2024-01-15","arxiv_id":"2401.07710","n_code_links":0,"syntology":null},{"paper":null,"title":"Curiosity-driven Exploration in Sparse-reward Multi-agent Reinforcement Learning","date":"2023-02-21","arxiv_id":"2302.10825","n_code_links":0,"syntology":null},{"paper":"/paper/time-myopic-go-explore-learning-a-state","title":"Time-Myopic Go-Explore: Learning A State Representation for the Go-Explore Paradigm","date":"2023-01-13","arxiv_id":"2301.05635","n_code_links":1,"syntology":{"ran":1,"of":3,"unverified":2,"pointer_only":0}},{"paper":null,"title":"First Go, then Post-Explore: the Benefits of Post-Exploration in Intrinsic Motivation","date":"2022-12-06","arxiv_id":"2212.03251","n_code_links":0,"syntology":null},{"paper":null,"title":"Quality-diversity in dissimilarity spaces","date":"2022-11-14","arxiv_id":"2211.12337","n_code_links":0,"syntology":null},{"paper":null,"title":"Go-Explore Complex 3D Game Environments for Automated Reachability Testing","date":"2022-09-01","arxiv_id":"2209.00570","n_code_links":0,"syntology":null},{"paper":"/paper/cell-free-latent-go-explore","title":"Cell-Free Latent Go-Explore","date":"2022-08-31","arxiv_id":"2208.14928","n_code_links":1,"syntology":null},{"paper":null,"title":"Generative Personas That Behave and Experience Like Humans","date":"2022-08-26","arxiv_id":"2209.00459","n_code_links":0,"syntology":null},{"paper":null,"title":"Play with Emotion: Affect-Driven Reinforcement Learning","date":"2022-08-26","arxiv_id":"2208.12622","n_code_links":0,"syntology":null},{"paper":null,"title":"When to Go, and When to Explore: The Benefit of Post-Exploration in Intrinsic Motivation","date":"2022-03-29","arxiv_id":"2203.16311","n_code_links":0,"syntology":null},{"paper":null,"title":"Go-Blend behavior and affect","date":"2021-09-24","arxiv_id":"2109.13388","n_code_links":0,"syntology":null},{"paper":"/paper/first-return-then-explore","title":"First return, then explore","date":"2020-04-27","arxiv_id":"2004.12919","n_code_links":2,"syntology":{"ran":2,"of":4,"unverified":2,"pointer_only":4}},{"paper":null,"title":"Adaptive Stress Testing without Domain Heuristics using Go-Explore","date":"2020-04-08","arxiv_id":"2004.04292","n_code_links":0,"syntology":null},{"paper":null,"title":"Exploration Based Language Learning for Text-Based Games","date":"2020-01-24","arxiv_id":"2001.08868","n_code_links":0,"syntology":null},{"paper":"/paper/go-explore-a-new-approach-for-hard","title":"Go-Explore: a New Approach for Hard-Exploration Problems","date":"2019-01-30","arxiv_id":"1901.10995","n_code_links":3,"syntology":{"ran":2,"of":4,"unverified":2,"pointer_only":4}}],"papers_shown":16,"tasks":[{"task":"/task/reinforcement-learning-1","name":"Reinforcement Learning (RL)","papers":8},{"task":"/task/reinforcement-learning","name":"Reinforcement Learning","papers":7},{"task":"/task/reinforcement-learning-2","name":"reinforcement-learning","papers":5},{"task":"/task/atari-games","name":"Atari Games","papers":3},{"task":"/task/montezumas-revenge","name":"Montezuma's Revenge","papers":3},{"task":"/task/imitation-learning","name":"Imitation Learning","papers":2},{"task":"/task/continuous-control","name":"Continuous Control","papers":1},{"task":"/task/decision-making","name":"Decision Making","papers":1},{"task":"/task/deep-reinforcement-learning","name":"Deep Reinforcement Learning","papers":1},{"task":"/task/diversity","name":"Diversity","papers":1},{"task":"/task/efficient-exploration","name":"Efficient Exploration","papers":1},{"task":"/task/management","name":"Management","papers":1},{"task":"/task/mathematical-reasoning","name":"Mathematical Reasoning","papers":1},{"task":"/task/mujoco","name":"MuJoCo","papers":1},{"task":"/task/multi-agent-reinforcement-learning","name":"Multi-agent Reinforcement Learning","papers":1},{"task":"/task/text-generation","name":"Text Generation","papers":1},{"task":"/task/continuous-control","name":"continuous-control","papers":1},{"task":"/task/energy-management","name":"energy management","papers":1},{"task":"/task/text-based-games","name":"text-based games","papers":1}],"tasks_shown":19,"n_tasks":19,"usage_by_year":[{"year":"2019","papers":1},{"year":"2020","papers":3},{"year":"2021","papers":1},{"year":"2022","papers":7},{"year":"2023","papers":2},{"year":"2024","papers":2}],"row_source":"methods_table","archive":{"source":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","archive_url":"https://paperswithcode.com/method/go-explore"},"syntology_read_at":"2026-09-24T18:15:14+00:00"}