{"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-to-search-with-mctsnets","title":"Learning to Search with MCTSnets","arxiv_id":"1802.04697","date":"2018-02-13","proceeding":"ICML 2018 7","authors":["Arthur Guez","Théophane Weber","Ioannis Antonoglou","Karen Simonyan","Oriol Vinyals","Daan Wierstra","Rémi Munos","David Silver"],"abstract":"Planning problems are among the most important and well-studied problems in\nartificial intelligence. They are most typically solved by tree search\nalgorithms that simulate ahead into the future, evaluate future states, and\nback-up those evaluations to the root of a search tree. Among these algorithms,\nMonte-Carlo tree search (MCTS) is one of the most general, powerful and widely\nused. A typical implementation of MCTS uses cleverly designed rules, optimized\nto the particular characteristics of the domain. These rules control where the\nsimulation traverses, what to evaluate in the states that are reached, and how\nto back-up those evaluations. In this paper we instead learn where, what and\nhow to search. Our architecture, which we call an MCTSnet, incorporates\nsimulation-based search inside a neural network, by expanding, evaluating and\nbacking-up a vector embedding. The parameters of the network are trained\nend-to-end using gradient-based optimisation. When applied to small searches in\nthe well known planning problem Sokoban, the learned search algorithm\nsignificantly outperformed MCTS baselines.","url_abs":"http://arxiv.org/abs/1802.04697v2","url_pdf":"http://arxiv.org/pdf/1802.04697v2.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-to-search-with-mctsnets","repo_url":"https://github.com/dixantmittal/mctsnet","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":{"status":"ok","spdx":"Apache-2.0"}},{"paper_slug":"learning-to-search-with-mctsnets","repo_url":"https://github.com/yashbonde/chess_lm","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"pytorch","reach":null}],"tasks":[{"task_slug":"sokoban","task_name":"Sokoban"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1802.04697","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}