{"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/expert-augmented-actor-critic-for-vizdoom-and","title":"Expert-augmented actor-critic for ViZDoom and Montezumas Revenge","arxiv_id":"1809.03447","date":"2018-09-10","proceeding":null,"authors":["Michał Garmulewicz","Henryk Michalewski","Piotr Miłoś"],"abstract":"We propose an expert-augmented actor-critic algorithm, which we evaluate on\ntwo environments with sparse rewards: Montezumas Revenge and a demanding maze\nfrom the ViZDoom suite. In the case of Montezumas Revenge, an agent trained\nwith our method achieves very good results consistently scoring above 27,000\npoints (in many experiments beating the first world). With an appropriate\nchoice of hyperparameters, our algorithm surpasses the performance of the\nexpert data. In a number of experiments, we have observed an unreported bug in\nMontezumas Revenge which allowed the agent to score more than 800,000 points.","url_abs":"http://arxiv.org/abs/1809.03447v1","url_pdf":"http://arxiv.org/pdf/1809.03447v1.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":"expert-augmented-actor-critic-for-vizdoom-and","repo_url":"https://github.com/ghostFaceKillah/expert","is_official":1,"mentioned_in_paper":1,"mentioned_in_github":1,"framework":"tf","reach":{"status":"unanswered"}},{"paper_slug":"expert-augmented-actor-critic-for-vizdoom-and","repo_url":"https://github.com/monte-ai/expert-augmented-acktr","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"tf","reach":{"status":"unanswered"}}],"tasks":[],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1809.03447","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}