让智能体用语言思考、交流、决策，提升经济策略的效率与鲁棒性。
Think, Speak, Decide: Language-Augmented Multi-Agent Reinforcement Learning for Economic Decision-Making
- 设计思考-说话-决策三阶段框架，融合语言与数值信息
- 在模拟经济中累计回报提升63.5%，鲁棒性提高18.8%
- 适合研究多智能体协作与语言增强决策的学者
经济决策不仅依赖价格、税收等结构化信号，还受非结构化语言（如同伴对话、媒体叙事）影响。尽管多智能体强化学习（MARL）在优化经济决策方面展现潜力，但难以处理语言的语义模糊性和上下文丰富性。本文提出语言增强多智能体策略（LAMP），采用“思考-说话-决策”流程：（1）思考阶段解析数值观测，提取短期冲击与长期趋势，并缓存高价值推理轨迹；（2）说话阶段基于推理生成并交换策略性消息，通过解析同伴沟通更新信念；（3）决策阶段融合数值数据、推理与反思，形成增强型MARL策略。实验表明，LAMP在经济模拟中相比MARL和仅使用大模型的基线，累计回报分别提升63.5%和34.0%，鲁棒性分别提高18.8%和59.4%，且可解释性更强。结果验证了语言增强策略在实现更有效、更稳健经济决策方面的潜力。
原文摘要 · Abstract (English)
Economic decision-making depends not only on structured signals such as prices and taxes, but also on unstructured language, including peer dialogue and media narratives. While multi-agent reinforcement learning (MARL) has shown promise in optimizing economic decisions, it struggles with the semantic ambiguity and contextual richness of language. We propose LAMP (Language-Augmented Multi-Agent Policy), a framework that integrates language into economic decision-making and narrows the gap to real-world settings. LAMP follows a Think-Speak-Decide pipeline: (1) Think interprets numerical observations to extract short-term shocks and long-term trends, caching high-value reasoning trajectories; (2) Speak crafts and exchanges strategic messages based on reasoning, updating beliefs by parsing peer communications; and (3) Decide fuses numerical data, reasoning, and reflections into a MARL policy to optimize language-augmented decision-making. Experiments in economic simulation show that LAMP outperforms both MARL and LLM-only baselines in cumulative return (+63.5%, +34.0%), robustness (+18.8%, +59.4%), and interpretability. These results demonstrate the potential of language-augmented policies to deliver more effective and robust economic strategies.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。