用分层树结构让大模型更高效完成复杂长任务
ReAcTree: Hierarchical LLM Agent Trees with Control Flow for Long-Horizon Task Planning
- 构建动态代理树，将大任务拆解为可管理的子目标
- 在WAH-NL上用Qwen 2.5 72B达61%成功率，是ReAct的两倍
- 适合需要多步规划的智能体任务，如机器人操作
大语言模型在具身智能体的决策与任务规划中取得显著进展，但现有方法在处理复杂长周期任务时受限于单一连贯路径，难以分离历史决策与观测。为此，我们提出ReAcTree，一种分层任务规划方法，通过动态构建代理树将复杂目标分解为可管理的子目标。每个子目标由具备推理、行动和扩展能力的LLM代理节点处理，控制流节点协调执行策略。同时，引入两种互补记忆系统：代理节点从情景记忆中检索特定目标的子目标示例，通过工作记忆共享环境观测。在WAH-NL和ALFRED数据集上的实验表明，ReAcTree持续优于强基准方法ReAct。尤其在WAH-NL上，使用Qwen 2.5 72B时达到61%的目标成功率，近乎是ReAct的31%的两倍。代码已开源。
原文摘要 · Abstract (English)
Recent advancements in large language models (LLMs) have enabled significant progress in decision-making and task planning for embodied autonomous agents. However, most existing methods struggle with complex, long-horizon tasks because they rely on a monolithic trajectory that entangles all past decisions and observations to solve the entire task in a single unified process. To address this limitation, we propose ReAcTree, a hierarchical task-planning method that decomposes a complex goal into manageable subgoals within a dynamically constructed agent tree. Each subgoal is handled by an LLM agent node capable of reasoning, acting, and further expanding the tree, while control flow nodes coordinate the execution strategies of agent nodes. In addition, we integrate two complementary memory systems: each agent node retrieves goal-specific, subgoal-level examples from episodic memory and shares environment-specific observations through working memory. Experiments on the WAH-NL and ALFRED show ReAcTree consistently outperforms strong task-planning baselines such as ReAct across diverse LLMs. Notably, on WAH-NL, ReAcTree achieves a 61% goal success rate with Qwen 2.5 72B, nearly doubling ReAct's 31%. The code is available at https://github.com/Choi-JaeWoo/ReAcTree.git.
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。