arXiv cs.MA· Mateusz Odrowaz-Sypniewski, Jasmine Bayrooti, Ajay Shankar, Amanda Prorok·· 11 小时前AI 评分22
多智能体强化学习中的广义意图建模
Generalized Intention Modeling in Multi-Agent Reinforcement Learning
AI 导读
研究提出一种任务自适应的对手建模框架，学习由性能驱动的多种意图表示混合，而非依赖先验选定的单一嵌入。该框架还引入一种新意图表示，最大化其与智能体自身未来回报的互信息，从而捕捉与性能最直接相关的对手信息。方法在多种任务上达到或超过 SOTA 基线，论文已被 NeurIPS 2026 接收。
正文 · 原文
Abstract:Modeling an opponent's intent is critical for effective decision-making in non-cooperative, competitive, and general-sum multi-agent reinforcement learning. Existing opponent modeling methods encode intent using an embedding derived from episode information chosen a priori, such as the opponent's next action or a future environment state, and use this to guide the ego-agent's behavior. These approaches assume that the chosen information is universally representative of intent; however, we show empirically that this is not the case as intentions are often task- and environment-dependent. To address this, we introduce a task-adaptive opponent modeling framework that learns a performance-driven mixture of multiple intent representations. We further introduce a new intention representation that maximizes mutual information with the ego-agent's future returns, thereby capturing opponent information that is most directly relevant to performance. Our approach consistently matches or exceeds the performance of state-of-the-art baselines across diverse tasks and yields insights into when and why different opponent modeling strategies succeed.
| Comments: | Accepted at NeurIPS 2026 |
| Subjects: | Machine Learning (cs.LG); Multiagent Systems (cs.MA) |
| Cite as: | arXiv:2605.31318 [cs.LG] |
| (or arXiv:2605.31318v3 [cs.LG] for this version) | |
| https://doi.org/10.48550/arXiv.2605.31318 arXiv-issued DOI via DataCite |
Submission history
From: Mateusz Odrowaz-Sypniewski [view email]
[v1]
Fri, 29 May 2026 13:48:18 UTC (2,897 KB)
[v2]
Fri, 30 Oct 2026 19:15:51 UTC (3,097 KB)
[v3]
Fri, 6 Nov 2026 13:35:40 UTC (3,104 KB)
来源：arXiv cs.MA · arxiv.org