Enhancing Spatial Reasoning in Large Language Models for Metal-Organic Frameworks Structure Prediction

2026-01-14

AI Summary

Accurately predicting the three-dimensional structures of metal–organic frameworks (MOFs) remains highly challenging due to their intricate atomic arrangements, which existing large language models struggle to represent effectively. This work proposes MOF-LLM, the first large language model framework tailored for modular MOF structure prediction, built upon Qwen-3 8B and enhanced with a block-level generation paradigm to improve spatial reasoning. The approach integrates spatial priors, continual pretraining (CPT), structure-supervised fine-tuning (SFT), and match-driven reinforcement learning, complemented by a novel Soft Adaptive Policy Optimization strategy to enhance structural stability. Experimental results demonstrate that MOF-LLM significantly outperforms current state-of-the-art denoising and language-model-based methods in both prediction accuracy and sampling efficiency.

Technology Category

• Machine Learning: Large Multimodal Models (LMMs)
• Natural Language Processing: (Large) Language Models
• Planning, Routing, and Scheduling: Planning with Language Models

Application Category

• User Modeling, Personalization and Recommendation: Large Language Models (LLM) for user modeling and recommendation
• Graph Algorithms and Modeling for the Web: Foundation models and LLMs for Web-related graphs
• Search and Retrieval-Augmented AI: Large language models for search

Abstract

Metal-organic frameworks (MOFs) are porous crystalline materials with broad applications such as carbon capture and drug delivery, yet accurately predicting their 3D structures remains a significant challenge. While Large Language Models (LLMs) have shown promise in generating crystals, their application to MOFs is hindered by MOFs'high atomic complexity. Inspired by the success of block-wise paradigms in deep generative models, we pioneer the use of LLMs in this domain by introducing MOF-LLM, the first LLM framework specifically adapted for block-level MOF structure prediction. To effectively harness LLMs for this modular assembly task, our training paradigm integrates spatial-aware continual pre-training (CPT), structural supervised fine-tuning (SFT), and matching-driven reinforcement learning (RL). By incorporating explicit spatial priors and optimizing structural stability via Soft Adaptive Policy Optimization (SAPO), our approach substantially enhances the spatial reasoning capability of a Qwen-3 8B model for accurate MOF structure prediction. Comprehensive experiments demonstrate that MOF-LLM outperforms state-of-the-art denoising-based and LLM-based methods while exhibiting superior sampling efficiency.

Problem

• Metal-Organic Frameworks
• Structure Prediction
• Spatial Reasoning
• Large Language Models
• 3D Structure

Innovation

• MOF-LLM
• spatial-aware continual pre-training
• block-level structure prediction
• Soft Adaptive Policy Optimization
• large language models for crystal generation