Beyond ReAct: A Planner-Centric Framework for Complex Tool-Augmented LLM Reasoning

TL;DR

This paper introduces a Planner-centric framework for large language models that improves complex query reasoning by global planning, surpassing existing methods like ReAct in multi-tool coordination and execution efficiency.

Contribution

The paper presents a novel Planner model with DAG-based global planning, a new benchmark dataset, and a two-stage training method to enhance tool coordination in LLM reasoning.

Findings

Achieves state-of-the-art performance on StableToolBench.

Demonstrates improved multi-tool coordination and complex query handling.

Outperforms ReAct in global planning and execution efficiency.

Abstract

Existing tool-augmented large language models (LLMs) encounter significant challenges when processing complex queries. Current frameworks such as ReAct are prone to local optimization traps due to their reliance on incremental decision-making processes. To address these limitations, we propose a novel Planner-centric Plan-Execute paradigm that fundamentally resolves local optimization bottlenecks through architectural innovation. Central to our approach is a novel Planner model that performs global Directed Acyclic Graph (DAG) planning for complex queries, enabling optimized execution beyond conventional tool coordination. We also introduce ComplexTool-Plan, a large-scale benchmark dataset featuring complex queries that demand sophisticated multi-tool composition and coordination capabilities. Additionally, we develop a two-stage training methodology that integrates Supervised Fine-Tuning (SFT) with Group Relative Policy Optimization (GRPO), systematically enhancing the Planner's tool selection accuracy and global planning awareness through structured DAG-based planning. When integrated with a capable executor, our framework achieves state-of-the-art performance on the StableToolBench benchmark for complex user queries, demonstrating superior end-to-end execution capabilities and robust handling of intricate multi-tool workflows.