Scholarlink

<p>Although RL agents are effective in well-defined environments, they often struggle to generalize to dynamic settings because the trajectories are typically gathered within a narrow training distribution. Under the potential distribution shift, policies overfit environment-specific cues, and the agent’s performance deteriorates. Recent work has explored applying Large Language Models (LLMs) or Vision Language Models (VLMs) to boost the generalization of RL agents through policy optimization guidance or prior knowledge. However, these approaches often lack seamless coordination between the RL agent and the foundation model, leading to unreasonable decision-making in unfamiliar environments and efficiency bottlenecks. Making full use of the inferential capabilities of foundation models and the rapid response capabilities of RL agents and enhancing the interaction between the two to form a dual system is still a lingering scientific question. To address this problem, we draw inspiration from Kahneman’s theory of fast thinking (System 1) and slow thinking (System 2), demonstrating that balancing intuition and deep reasoning can achieve nimble decision-making in a complex world. In this study, we propose a <b>Dual-System Adaptive Decision Framework</b> (DSADF), integrating two complementary modules: System 1, comprising an RL agent and a memory space for fast and intuitive decision making, and System 2, driven by a VLM for deep and analytical reasoning. DSADF facilitates efficient and adaptive decision-making by combining the strengths of both systems. The empirical study in the video game environment: <i>Crafter</i>, <i>Housekeep</i>, and <i>AI2-THOR</i> demonstrates the effectiveness of our proposed method, showing significant improvements in decision abilities for both unseen and known tasks.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

DSADF: Thinking Fast and Slow for Decision Making

  • 期刊论文

  • 发表日期：

    2026年5月14日

查看全文
  • Zhihao Dou,
  • Dongfei Cui,
  • Jun Yan,
  • Weida Wang,
  • Benteng Chen,
  • Haoming Wang,
  • Zeke Xie,
  • Shufei Zhang

摘要

Although RL agents are effective in well-defined environments, they often struggle to generalize to dynamic settings because the trajectories are typically gathered within a narrow training distribution. Under the potential distribution shift, policies overfit environment-specific cues, and the agent’s performance deteriorates. Recent work has explored applying Large Language Models (LLMs) or Vision Language Models (VLMs) to boost the generalization of RL agents through policy optimization guidance or prior knowledge. However, these approaches often lack seamless coordination between the RL agent and the foundation model, leading to unreasonable decision-making in unfamiliar environments and efficiency bottlenecks. Making full use of the inferential capabilities of foundation models and the rapid response capabilities of RL agents and enhancing the interaction between the two to form a dual system is still a lingering scientific question. To address this problem, we draw inspiration from Kahneman’s theory of fast thinking (System 1) and slow thinking (System 2), demonstrating that balancing intuition and deep reasoning can achieve nimble decision-making in a complex world. In this study, we propose a Dual-System Adaptive Decision Framework (DSADF), integrating two complementary modules: System 1, comprising an RL agent and a memory space for fast and intuitive decision making, and System 2, driven by a VLM for deep and analytical reasoning. DSADF facilitates efficient and adaptive decision-making by combining the strengths of both systems. The empirical study in the video game environment: Crafter, Housekeep, and AI2-THOR demonstrates the effectiveness of our proposed method, showing significant improvements in decision abilities for both unseen and known tasks.