arXiv cs.MA· Li Ma, Hao Peng, Yiming Wang, Jie Liu, Kongjing Gu, Kangli Zi, Na Han, Hui Lin·· 9 小时前AI 评分31
SEMA:面向 RTS 场景高效决策的自进化多智能体框架
Self-Evolving Multi-Agent Framework for Efficient Decision Making in Real-Time Strategy Scenarios
AI 导读
研究者提出自进化多智能体框架 SEMA,用于实时策略(RTS)场景下的高性能低延迟决策。该框架通过回合内评估与跨回合分析自适应校准模型偏差,并基于结构熵的动态观测剪枝将高维状态蒸馏为核心语义信息。在多个 StarCraft II 地图上的实验显示,SEMA 取得更高胜率的同时将平均决策延迟降低 50% 以上。
正文 · 原文
Abstract:Large language models (LLMs) have demonstrated exceptional potential in complex reasoning,pioneering a new paradigm for autonomous agent decision making in dynamic settings. However, in Real-Time Strategy (RTS) scenarios, LLMs suffer from a critical speed-quality trade-off. Specifically expansive state spaces and time limits render inference delays prohibitive, while stochastic planning errors undermine logical consistency. To address these challenges, we present SEMA (Self-Evolving Multi-Agent), a novel framework designed for high-performance, low-latency decision-making in RTS environments. This collaborative multi-agent framework facilitates self-evolution by adaptively calibrating model bias through in-episode assessment and cross-episode analysis. We further incorporate dynamic observation pruning based on structural entropy to model game states topologically. By distilling high dimensional data into core semantic information, this approach significantly reduces inference time. We also develop a hybrid knowledge-memory mechanism that integrates micro-trajectories, macro-experience, and hierarchical domain knowledge, thereby enhancing both strategic adaptability and decision consistency. Experiments across multiple StarCraft II maps demonstrate that SEMA achieves superior win rates while reducing average decision latency by over 50%, validating its efficiency and robustness in complex RTS scenarios.
| Comments: | 17 pages, 6 figures |
| Subjects: | Multiagent Systems (cs.MA) |
| Cite as: | arXiv:2603.23875 [cs.MA] |
| (or arXiv:2603.23875v2 [cs.MA] for this version) | |
| https://doi.org/10.48550/arXiv.2603.23875 arXiv-issued DOI via DataCite |
Submission history
From: Yiming Wang [view email]
[v1]
Wed, 25 Mar 2026 03:05:29 UTC (985 KB)
[v2]
Thu, 19 Nov 2026 15:27:08 UTC (987 KB)
来源:arXiv cs.MA · arxiv.org