{"type":"ScholarlyArticle","title":"Joint-Perturbation Simultaneous Pseudo-Gradient","authors":[{"name":"Carlos Martin","slug":"carlos-martin","given":"Carlos","family":"Martin","affiliation":""},{"name":"Tuomas Sandholm","slug":"tuomas-sandholm","given":"Tuomas","family":"Sandholm","affiliation":""}],"abstract":"We study the problem of computing an approximate Nash equilibrium of a game whose strategy space is continuous without access to gradients of the utility function.\r\nLack of access to gradients is common in reinforcement learning settings, where the environment is treated as a black box, as well as equilibrium finding in mechanisms such as auctions, where the mechanism's payoffs are discontinuous in the players' actions.\r\nTo tackle this problem, we turn to zeroth-order optimization techniques that combine pseudo-gradients with equilibrium-finding dynamics.\r\nSpecifically, we introduce a new technique that requires a number of utility function evaluations per iteration that is constant rather than linear in the number of players.\r\nIt achieves this by performing a single joint perturbation on all players' strategies, rather than perturbing each one individually.\r\nThis is very important for many-player games, especially when the utility function is expensive to compute in terms of wall time, memory, money, or other resources.\r\nWe evaluate our approach on various games, including auctions, which have important real-world applications.\r\nOur approach yields a dramatic improvement in performance in terms of the wall time required to reach an approximate Nash equilibrium.","doi":"10.24963/ijcai.2025/445","url":"https://ijcai.org/proceedings/2025/445","pdf":"https://ijcai.org/proceedings/2025/0445.pdf","markdown":"https://ijcai.org/proceedings/2025/445.md","bibtex":"https://ijcai.org/proceedings/2025/bibtex/445","volume":{"slug":"2025","title":"Proceedings of the Thirty-Fourth International Joint Conference on Artificial Intelligence","short_title":"IJCAI 2025","url":"https://ijcai.org/proceedings/2025/","year":2025,"location":"Montreal, Canada","dates":"16-22 August 2025","editors":"James Kwok","isbn":"978-1-956792-06-5","publisher":"International Joint Conferences on Artificial Intelligence Organization","doi_prefix":"10.24963/ijcai.2025"},"section":"Game Theory and Economic Paradigms","track":"Main Track","number":445,"pages":{"first":3996,"last":4004},"keywords":["Game Theory and Economic Paradigms: GTEP: Noncooperative games"],"videos":[],"references":[],"license_note":"Copyright International Joint Conferences on Artificial Intelligence."}