{"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/a-deeper-look-at-experience-replay","title":"A Deeper Look at Experience Replay","arxiv_id":"1712.01275","date":"2017-12-04","proceeding":null,"authors":["Shangtong Zhang","Richard S. Sutton"],"abstract":"Recently experience replay is widely used in various deep reinforcement\nlearning (RL) algorithms, in this paper we rethink the utility of experience\nreplay. It introduces a new hyper-parameter, the memory buffer size, which\nneeds carefully tuning. However unfortunately the importance of this new\nhyper-parameter has been underestimated in the community for a long time. In\nthis paper we did a systematic empirical study of experience replay under\nvarious function representations. We showcase that a large replay buffer can\nsignificantly hurt the performance. Moreover, we propose a simple O(1) method\nto remedy the negative influence of a large replay buffer. We showcase its\nutility in both simple grid world and challenging domains like Atari games.","url_abs":"http://arxiv.org/abs/1712.01275v3","url_pdf":"http://arxiv.org/pdf/1712.01275v3.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":"a-deeper-look-at-experience-replay","repo_url":"https://github.com/VictorZuanazzi/Project_RL","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":null},{"paper_slug":"a-deeper-look-at-experience-replay","repo_url":"https://github.com/danielytan/doom_DQN","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"unanswered"}},{"paper_slug":"a-deeper-look-at-experience-replay","repo_url":"https://github.com/rikluost/RL_DQN_Pong","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":{"status":"unanswered"}},{"paper_slug":"a-deeper-look-at-experience-replay","repo_url":"https://github.com/seungjaeryanlee/combined-experience-replay","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"unanswered"}}],"tasks":[{"task_slug":"atari-games","task_name":"Atari Games"},{"task_slug":"deep-reinforcement-learning","task_name":"Deep Reinforcement Learning"},{"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":[{"method_slug":"experience-replay","method_name":"Experience Replay"}],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"syntology_url":"https://syntology.ai/paper/1712.01275","atlas_url":"https://app.syntology.ai/?focus=1712.01275","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}