GATS用分层世界模型替代LLM推理，实现零调用、百分百成功的高效规划。
GATS: Graph-Augmented Tree Search with Layered World Models for Efficient Agent Planning
- 用三层世界模型（符号匹配/日志统计/大模型预测）替代规划中频繁调用LLM
- 在12个复杂场景中达成100%成功率，远超LATS的88.9%和ReAct的23.9%
- 规划全程无需调用LLM，结果完全确定，适合对稳定性和效率要求高的任务
大型语言模型（LLM）代理在多步规划任务中展现出潜力，但现有方法如LATS（语言代理树搜索）和ReAct严重依赖规划过程中的LLM推理，导致计算成本高且行为随机。我们提出 extbf{GATS}（图增强树搜索），一种结合系统化UCB1树搜索与分层世界模型的规划框架，在推理阶段完全消除对LLM的调用，同时实现更优的规划性能。我们的三层世界模型包括：（L1）精确的符号动作匹配，（L2）从执行日志中学习的统计信息，（L3）对未知动作的LLM预测。在具有分支路径和死胡同的合成规划任务中，GATS达到100%成功率，优于LATS的92%和ReAct的64%。在涵盖12个挑战性场景的综合压力测试中——包括编码工作流、网页导航和长程任务——GATS维持100%成功率，而LATS降至88.9%，ReAct降至23.9%。GATS在规划阶段每任务仅需0次LLM调用（相比LATS的37次），并生成无方差的确定性计划。结果表明，结合学习的世界模型与系统性搜索可显著超越基于LLM引导的探索用于代理规划。
原文摘要 · Abstract (English)
Large Language Model (LLM) agents have shown promise in multi-step planning tasks, but existing approaches like LATS (Language Agent Tree Search) and ReAct rely heavily on LLM inference during planning, leading to high computational costs and stochastic behavior. We present \textbf{GATS} (Graph-Augmented Tree Search), a planning framework that combines systematic UCB1-based tree search with a layered world model to eliminate LLM calls during inference while achieving superior planning performance. Our three-layer world model integrates: (L1) exact symbolic action matching, (L2) statistics learned from execution logs, and (L3) LLM-based prediction for unknown actions. On synthetic planning tasks with branching paths and dead-ends, GATS achieves \textbf{100\% success rate} compared to 92 % for LATS and 64\% for ReAct. On a comprehensive stress test spanning 12 challenging scenarios -- including coding workflows, web navigation, and long-horizon tasks -- GATS maintains \textbf{100\% success} while LATS drops to 88.9 % and ReAct to 23.9%. GATS requires \textbf{zero LLM calls per task} during planning (vs. 37 per task for LATS) and produces deterministic plans with zero variance across runs. Our results demonstrate that systematic search with learned world models can substantially outperform LLM-guided exploration for agent planning.
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。