ReAcTree: Hierarchical LLM Agent Trees with Control Flow for Long-Horizon Task Planning

Jae-Woo Choi, Hyungmin Kim, Hyobin Ong, Youngwoo Yoon, Minsu Jang, DohyungKim, Jaehong Kim

Keywords: Task planning, large language models, decision-making, hierarchical planning, behavior trees, embodied agents

Abstract: 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 still struggle with complex, long-horizon tasks because they rely on a monolithic trajectory that entangles all past decisions and observations, attempting to solve the entire task in a single unified process. To address this limitation, we propose $\textit{ReAcTree}$, a hierarchical task-planning method that decomposes a complex goal into more 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 $\textit{episodic memory}$ and shares environment-specific observations through $\textit{working memory}$. Experiments on the WAH-NL and ALFRED datasets demonstrate that $\textit{ReAcTree}$ consistently outperforms strong task-planning baselines such as $\textit{ReAct}$ across diverse LLMs. Notably, on WAH-NL, $\textit{ReAcTree}$ achieves a 61\% goal success rate with Qwen 2.5 72B, nearly doubling $\textit{ReAct}$’s 31\%.

Area: Generative and Agentic AI (GAAI)

Generative A I: I acknowledge that I have read and will follow this policy.

Submission Number: 190