MEIDNet: Multimodal generative AI framework for inverse materials design

2026-01-29

AI Summary

This work addresses the challenge of jointly optimizing structure and multiple properties in inverse design of novel materials by proposing MEIDNet, a framework that, for the first time, integrates structural, electronic, and optical modalities to construct a highly aligned latent space. By combining equivariant graph neural networks, multimodal contrastive learning, and generative inverse design—augmented with a curriculum learning strategy—the method achieves a ~60-fold improvement in training efficiency and significantly enhances generation quality. Experimental results demonstrate a latent space cosine similarity of 0.96 and successful generation of a high proportion of stable, novel, and unique low-bandgap perovskite structures, achieving a SUN rate of 13.6%, with validity confirmed by ab initio calculations.

Technology Category

• Machine Learning: Multimodal Learning
• Search and Optimization: Learning to Search
• Computer Vision: Generative Adversarial Networks (GANs) for Vision

Application Category

• Graph Algorithms and Modeling for the Web: Graph embeddings and representation learning for Web-related graphs
• Economics, Online Markets and Human Computation: Economic ramifications for generative AI infrastructure and applications
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

In this work, we present Multimodal Equivariant Inverse Design Network (MEIDNet), a framework that jointly learns structural information and materials properties through contrastive learning, while encoding structures via an equivariant graph neural network (EGNN). By combining generative inverse design with multimodal learning, our approach accelerates the exploration of chemical-structural space and facilitates the discovery of materials that satisfy predefined property targets. MEIDNet exhibits strong latent-space alignment with cosine similarity 0.96 by fusion of three modalities through cross-modal learning. Through implementation of curriculum learning strategies, MEIDNet achieves ~60 times higher learning efficiency than conventional training techniques. The potential of our multimodal approach is demonstrated by generating low-bandgap perovskite structures at a stable, unique, and novel (SUN) rate of 13.6 %, which are further validated by ab initio methods. Our inverse design framework demonstrates both scalability and adaptability, paving the way for the universal learning of chemical space across diverse modalities.

Problem

• inverse materials design
• multimodal learning
• chemical-structural space
• generative AI
• property-targeted discovery

Innovation

• multimodal learning
• equivariant graph neural network
• inverse materials design
• contrastive learning
• curriculum learning