{"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/playing-atari-games-with-deep-reinforcement","title":"Playing Atari Games with Deep Reinforcement Learning and Human Checkpoint Replay","arxiv_id":"1607.05077","date":"2016-07-18","proceeding":null,"authors":["Ionel-Alexandru Hosu","Traian Rebedea"],"abstract":"This paper introduces a novel method for learning how to play the most\ndifficult Atari 2600 games from the Arcade Learning Environment using deep\nreinforcement learning. The proposed method, human checkpoint replay, consists\nin using checkpoints sampled from human gameplay as starting points for the\nlearning process. This is meant to compensate for the difficulties of current\nexploration strategies, such as epsilon-greedy, to find successful control\npolicies in games with sparse rewards. Like other deep reinforcement learning\narchitectures, our model uses a convolutional neural network that receives only\nraw pixel inputs to estimate the state value function. We tested our method on\nMontezuma's Revenge and Private Eye, two of the most challenging games from the\nAtari platform. The results we obtained show a substantial improvement compared\nto previous learning approaches, as well as over a random player. We also\npropose a method for training deep reinforcement learning agents using human\ngameplay experience, which we call human experience replay.","url_abs":"http://arxiv.org/abs/1607.05077v1","url_pdf":"http://arxiv.org/pdf/1607.05077v1.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":"playing-atari-games-with-deep-reinforcement","repo_url":"https://github.com/smj007/Breakout_A3C","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok"}}],"tasks":[{"task_slug":"atari-games","task_name":"Atari Games"},{"task_slug":"deep-reinforcement-learning","task_name":"Deep Reinforcement Learning"},{"task_slug":"montezumas-revenge","task_name":"Montezuma's Revenge"},{"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=1607.05077","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}