Model Agnostic Graph Prompt Learning for Crystal Property Prediction

2026-07-09

AI Summary

This work addresses the limitations of existing graph neural networks (GNNs) in crystal property prediction, which often rely heavily on domain-specific knowledge, involve large parameter counts, and struggle to comprehensively model essential chemical and structural features. To overcome these challenges, the authors propose a lightweight, model-agnostic, multi-level graph prompt learning framework that introduces soft prompting into crystal graph representation learning for the first time. Specifically, node-level soft prompts capture local atomic chemical semantics, while graph-level soft prompts encode global structural symmetries. This approach enhances the backbone GNN’s capacity to learn latent features without altering its architecture. The framework facilitates cross-property knowledge transfer, consistently improving the performance of state-of-the-art GNNs by 3%–15% across multiple benchmark datasets and significantly boosting generalization under data-scarce conditions.

Abstract

Graph Neural Networks have emerged as a powerful tool for the fast and accurate prediction of various crystal properties. These models often encode domain-specific knowledge into their graph encoding modules, which increases their parameter size and makes their performance heavily dependent on domain expertise. Added to this, explicitly incorporating all chemical and structural features, that might influence a specific crystal property into the GNN encoder, is a challenging task. In this work, we propose a soft prompt learning framework that captures latent features essential for property prediction, which are not explicitly provided to the GNN. We introduce a novel multilevel graph prompt learning framework comprising both node-level and graph-level soft prompts. At the node level, we capture the local chemical semantics of different atom types, while at the graph level, we encode the global structural symmetry of the crystal graph. Our proposed prompt learning framework is lightweight and seamlessly integrates with any existing GNN encoder. Extensive experiments on popular benchmark datasets show that incorporating prompt learning significantly improves (3\% - 15\%) the performance of state-of-the-art GNN models in crystal property prediction tasks. Furthermore, the learned soft prompts enable cross-property knowledge transfer, enhancing prediction performance for properties with limited training data. Code is available at https://github.com/shrimonmuke0202/Prompt.git

Problem

• Crystal Property Prediction
• Graph Neural Networks
• Domain Knowledge
• Feature Encoding
• Model Agnosticism

Innovation

• graph prompt learning
• model-agnostic
• crystal property prediction
• soft prompts
• graph neural networks