Saved in:
Bibliographic Details
Main Authors: Chakraborty, Sourav, Rege, Amit Kiran, Monteleoni, Claire, Chen, Lijun
Format: Preprint
Published: 2026
Subjects:
Online Access:https://arxiv.org/abs/2602.16965
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866914338677194752
author Chakraborty, Sourav
Rege, Amit Kiran
Monteleoni, Claire
Chen, Lijun
author_facet Chakraborty, Sourav
Rege, Amit Kiran
Monteleoni, Claire
Chen, Lijun
contents We study the decentralized multi-player stochastic bandit problem over a continuous, Lipschitz-structured action space where hard collisions yield zero reward. Our objective is to design a communication-free policy that maximizes collective reward, with coordination costs that are independent of the time horizon $T$. We propose a modular protocol that first solves the multi-agent coordination problem -- identifying and seating players on distinct high-value regions via a novel maxima-directed search -- and then decouples the problem into $N$ independent single-player Lipschitz bandits. We establish a near-optimal regret bound of $\tilde{O}(T^{(d+1)/(d+2)})$ plus a $T$-independent coordination cost, matching the single-player rate. To our knowledge, this is the first framework providing such guarantees, and it extends to general distance-threshold collision models.
format Preprint
id arxiv_https___arxiv_org_abs_2602_16965
institution arXiv
publishDate 2026
record_format arxiv
spellingShingle Multi-Agent Lipschitz Bandits
Chakraborty, Sourav
Rege, Amit Kiran
Monteleoni, Claire
Chen, Lijun
Machine Learning
We study the decentralized multi-player stochastic bandit problem over a continuous, Lipschitz-structured action space where hard collisions yield zero reward. Our objective is to design a communication-free policy that maximizes collective reward, with coordination costs that are independent of the time horizon $T$. We propose a modular protocol that first solves the multi-agent coordination problem -- identifying and seating players on distinct high-value regions via a novel maxima-directed search -- and then decouples the problem into $N$ independent single-player Lipschitz bandits. We establish a near-optimal regret bound of $\tilde{O}(T^{(d+1)/(d+2)})$ plus a $T$-independent coordination cost, matching the single-player rate. To our knowledge, this is the first framework providing such guarantees, and it extends to general distance-threshold collision models.
title Multi-Agent Lipschitz Bandits
topic Machine Learning
url https://arxiv.org/abs/2602.16965