ARMOR: An Agentic Framework for Reaction Feasibility Prediction via Adaptive Utility-aware Multi-tool Reasoning

2026-05-07

AI Summary

This work addresses the challenge of substantial performance discrepancies and frequent conflicts among multiple tools in chemical reaction feasibility prediction by proposing a large language model–based agent framework. The framework enables adaptive collaborative reasoning through hierarchical tool organization, tool-specific utility modeling, and a memory-augmented conflict resolution mechanism. It dynamically assesses tool reliability and preferentially invokes high-confidence tools during inference. Evaluated on public benchmarks, the proposed method significantly outperforms existing single-tool and multi-tool fusion approaches, demonstrating particularly strong performance in scenarios where tool predictions conflict. This leads to notable improvements in both accuracy and robustness of reaction feasibility prediction.

Abstract

Reaction feasibility prediction, as a fundamental problem in computational chemistry, has benefited from diverse tools enabled by recent advances in artificial intelligence, particularly large language models. However, the performance of individual tools varies substantially across reactions, making it difficult for any single tool to consistently perform well across all cases. This raises a critical challenge: how to effectively leverage multiple tools to obtain more accurate feasibility predictions. To address this, we propose ARMOR, an agentic framework that explicitly models tool-specific utilities, adaptively prioritizes tools, and further resolves the potential tool conflicts to produce the final prediction for each reaction. Unlike existing approaches that rely on simple aggregation or heuristic assignment over various tools, ARMOR organizes tools into a hierarchy that prioritizes top-performing tools and defers others when needed, characterizes their strengths through tool-specific patterns, and resolves conflicts via memoryaugmented reasoning. Extensive experiments on a public dataset demonstrate that ARMOR consistently outperforms strong baselines, including single-tool methods as well as various tool aggregation and tool selection approaches. Further analysis shows that the improvements are particularly significant on reactions with conflicting tool predictions, highlighting the effectiveness of ARMOR in leveraging the complementary strengths of multiple tools. The code is available via https://anonymous.4open.science/r/ARMOR-E13F.

Problem

• reaction feasibility prediction
• computational chemistry
• tool integration
• multi-tool reasoning
• AI in chemistry

Innovation

• agentic framework
• reaction feasibility prediction
• adaptive tool selection
• utility-aware reasoning
• memory-augmented conflict resolution