{"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/learning-actionable-representations-with-goal","title":"Learning Actionable Representations with Goal-Conditioned Policies","arxiv_id":"1811.07819","date":"2018-11-19","proceeding":null,"authors":["Dibya Ghosh","Abhishek Gupta","Sergey Levine"],"abstract":"Representation learning is a central challenge across a range of machine\nlearning areas. In reinforcement learning, effective and functional\nrepresentations have the potential to tremendously accelerate learning progress\nand solve more challenging problems. Most prior work on representation learning\nhas focused on generative approaches, learning representations that capture all\nunderlying factors of variation in the observation space in a more disentangled\nor well-ordered manner. In this paper, we instead aim to learn functionally\nsalient representations: representations that are not necessarily complete in\nterms of capturing all factors of variation in the observation space, but\nrather aim to capture those factors of variation that are important for\ndecision making -- that are \"actionable.\" These representations are aware of\nthe dynamics of the environment, and capture only the elements of the\nobservation that are necessary for decision making rather than all factors of\nvariation, without explicit reconstruction of the observation. We show how\nthese representations can be useful to improve exploration for sparse reward\nproblems, to enable long horizon hierarchical reinforcement learning, and as a\nstate representation for learning policies for downstream tasks. We evaluate\nour method on a number of simulated environments, and compare it to prior\nmethods for representation learning, exploration, and hierarchical\nreinforcement learning.","url_abs":"http://arxiv.org/abs/1811.07819v2","url_pdf":"http://arxiv.org/pdf/1811.07819v2.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":"learning-actionable-representations-with-goal","repo_url":"https://github.com/elitalobo/Hierarchical-RL-Algorithms","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":null}],"tasks":[{"task_slug":"decision-making","task_name":"Decision Making"},{"task_slug":"hierarchical-reinforcement-learning","task_name":"Hierarchical Reinforcement Learning"},{"task_slug":"reinforcement-learning","task_name":"Reinforcement Learning"},{"task_slug":"reinforcement-learning-1","task_name":"Reinforcement Learning (RL)"},{"task_slug":"representation-learning","task_name":"Representation Learning"},{"task_slug":"reinforcement-learning-2","task_name":"reinforcement-learning"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1811.07819","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}