Hanabi 通用智能体 R3D2 论文因实现 bug 被撤稿

This paper has been withdrawn by Hadi Nekoei

No PDF available, click to view other formats

Abstract:Traditional multi-agent reinforcement learning (MARL) systems can develop cooperative strategies through repeated interactions. However, these systems are unable to perform well on any other setting than the one they have been trained on, and struggle to successfully cooperate with unfamiliar collaborators. This is particularly visible in the Hanabi benchmark, a popular 2-to-5 player cooperative card-game which requires complex reasoning and precise assistance to other agents. Current MARL agents for Hanabi can only learn one specific game-setting (e.g., 2-player games), and play with the same algorithmic agents. This is in stark contrast to humans, who can quickly adjust their strategies to work with unfamiliar partners or situations. In this paper, we introduce Recurrent Replay Relevance Distributed DQN (R3D2), a generalist agent for Hanabi, designed to overcome these limitations. We reformulate the task using text, as language has been shown to improve transfer. We then propose a distributed MARL algorithm that copes with the resulting dynamic observation- and action-space. In doing so, our agent is the first that can play all game settings concurrently, and extend strategies learned from one setting to other ones. As a consequence, our agent also demonstrates the ability to collaborate with different algorithmic agents -- agents that are themselves unable to do so. The implementation code is available at: $\href{this https URL}{R3D2-A-Generalist-Hanabi-Agent}$

Comments:	This paper has been withdrawn because we found a bug in R3D2 implementation that invalidates some of the key claims in the paper
Subjects:	Multiagent Systems (cs.MA); Artificial Intelligence (cs.AI)
Cite as:	arXiv:2503.14555 [cs.MA]
	(or arXiv:2503.14555v2 [cs.MA] for this version)
	https://doi.org/10.48550/arXiv.2503.14555 arXiv-issued DOI via DataCite

Submission history

From: Hadi Nekoei [view email]
[v1] Mon, 17 Mar 2025 22:25:15 UTC (1,631 KB)
[v2] Fri, 9 Oct 2026 19:39:08 UTC (1 KB) (withdrawn)