MOF-Sleuth: Tool-Grounded Reward Alignment for Explainable Fine-Grained MOF CIF Auditing

2026-07-22

AI Summary

This work addresses the prevalence of subtle chemical and structural errors in metal–organic framework (MOF) Crystallographic Information Files (CIFs), which critically undermine the reliability of downstream simulations and machine learning applications, while existing validation methods lack interpretable, fine-grained diagnostic capabilities. The authors propose a reinforcement learning–driven CIF auditing agent that integrates deterministic chemical computation modules—encompassing composition, geometry, connectivity, site occupancy, coordination, and charge—with a large language model reasoning engine to produce interpretable error diagnoses and binary validity judgments grounded in tool-generated evidence. A novel reward-guided mechanism translates measurement signals into chemically meaningful supervision, and a newly introduced Chem-GD metric evaluates the consistency between diagnoses and supporting evidence. Evaluated across four benchmarks, the method significantly outperforms both general large language models and MOF-specific approaches in error detection, attribution accuracy, and explanation quality.

Abstract

Large metal-organic framework (MOF) databases support simulation, screening, and machine learning through crystallographic information files (CIFs). Subtle chemical and structural errors in these inputs can compromise downstream results and hinder manual inspection. LLM advances in computational chemistry offer paths beyond predictive screening toward fine-grained diagnosis with evidence-grounded explanations. However, two challenges remain: (i) limited fine-grained attribution: MOF-specific validators and machine-learning models scale detection but provide fixed checks, readiness scores, or coarse labels rather than evidence-grounded explanations; and (ii) unreliable CIF reasoning: direct LLM auditing is costly and unreliable because chemical evidence is implicit across atom-site records and requires geometric, connectivity, occupancy, and charge calculations. Both stem from weak coupling between chemical evidence and language-model explanation. We introduce MOF-Sleuth, a reinforcement-guided CIF auditing agent with two modules: a deterministic Forensic Lab and a Sleuth reasoning engine. The Lab derives composition, geometry, connectivity, occupancy, coordination, and charge evidence, and Sleuth uses this evidence to produce an evidence-grounded explanation, error types, and a binary decision. Reward-guided reinforcement learning (RL) turns tool measurements into chemical explanation-level supervision, rewarding not only the final answer but also cited chemical evidence and evidence-supported diagnoses. We introduce Chemically Grounded Diagnosis (Chem-GD), a metric that assesses whether a correct diagnosis is explained by factual, relevant CIF-derived evidence. Across four benchmarks, MOF-Sleuth establishes state-of-the-art performance among LLM-based approaches and MOF-specific machine-learning methods, demonstrating gains in detection, attribution, and grounded explanation quality.

Problem

• MOF
• CIF auditing
• explainable diagnosis
• chemical evidence
• fine-grained error detection

Innovation

• MOF-Sleuth
• reward-guided reinforcement learning
• evidence-grounded explanation
• Chem-GD
• CIF auditing