ChemHTS: Hierarchical Tool Stacking for Enhancing Chemical Agents

TL;DR

ChemHTS introduces a hierarchical tool stacking approach that improves the collaboration and accuracy of LLMs in chemistry tasks by optimizing tool invocation pathways and decision-making processes.

Contribution

It proposes a novel hierarchical stacking strategy for tool invocation in LLMs, enhancing performance and interpretability in chemical research applications.

Findings

Outperforms baseline models like GPT-4o and ChemDFM on chemistry tasks.

Demonstrates improved tool collaboration and decision accuracy.

Provides insights into tool-stacking behaviors for better interpretability.

Abstract

Large Language Models (LLMs) have demonstrated remarkable potential in scientific research, particularly in chemistry-related tasks such as molecular design, reaction prediction, and property estimation. While tool-augmented LLMs have been introduced to enhance reasoning and computation in these domains, existing approaches suffer from tool invocation errors and lack effective collaboration among diverse tools, limiting their overall performance. To address these challenges, we propose ChemHTS (Chemical Hierarchical Tool Stacking), a novel method that optimizes tool invocation pathways through a hierarchical stacking strategy. ChemHTS consists of two key stages: tool self-stacking warmup and multi-layer decision optimization, enabling LLMs to refine tool usage dynamically. We evaluate ChemHTS across four classical chemistry tasks and demonstrate its superiority over strong baselines, including GPT-4o, DeepSeek-R1, and chemistry-specific models, including ChemDFM. Furthermore, we define four distinct tool-stacking behaviors to enhance interpretability, providing insights into the effectiveness of tool collaboration. Our dataset and code are publicly available at \url{https://github.com/Chang-pw/ChemHTS}.