[2506.10127] Meet Me at the Arm: The Cooperative Multi-Armed Bandits Problem with Shareable Arms

Computer Science > Machine Learning arXiv:2506.10127 (cs) [Submitted on 11 Jun 2025 (v1), last revised 29 Mar 2026 (this version, v2)] Title:Meet Me at the Arm: The Cooperative Multi-Armed Bandits Problem with Shareable Arms Authors:Xinyi Hu, Aldo Pacchiano View a PDF of the paper titled Meet Me at the Arm: The Cooperative Multi-Armed Bandits Problem with Shareable Arms, by Xinyi Hu and Aldo Pacchiano View PDF HTML (experimental) Abstract:We study the decentralized multi-player multi-armed bandits (MMAB) problem under a no-sensing setting, where each player receives only their own reward and obtains no information about collisions. Each arm has an unknown capacity, and if the number of players pulling an arm exceeds its capacity, all players involved receive zero reward. This setting generalizes the classical unit-capacity model and introduces new challenges in coordination and capacity discovery under severe feedback limitations. We propose A-CAPELLA (Algorithm for Capacity-Aware Parallel Elimination for Learning and Allocation), a decentralized learning algorithm that achieves logarithmic regret in this generalized regime via protocol-driven coordination. Subjects: Machine Learning (cs.LG) Cite as: arXiv:2506.10127 [cs.LG]   (or arXiv:2506.10127v2 [cs.LG] for this version)   https://doi.org/10.48550/arXiv.2506.10127 Focus to learn more arXiv-issued DOI via DataCite Submission history From: Xinyi Hu [view email] [v1] Wed, 11 Jun 2025 19:14:32 UTC (5,951 KB) [v2] Sun, 29 M...