Publications

Conference Papers

Accurately Predicting Protein Mutational Effects via a Hierarchical Many-Body Attention Network

Published in Annual Conference on Neural Information Processing Systems (NeurIPS), 2025

We propose H3-DDG, a Hypergraph-driven Hierarchical network to capture Higher-order many-body interactions across multiple scales. By introducing a hierarchical communication mechanism, H3-DDG effectively models both local and global mutational effects.

Download Paper | Download Bibtex

Quadruple Attention in Many-body Systems for Accurate Molecular Property Predictions

Published in International Conference on Machine Learning (ICML), 2025

MABNet introduces a geometric attention framework that efficiently models four-body atomic interactions, overcoming limitations of traditional methods in capturing complex many-body effects, and achieves state-of-the-art performance on molecular property prediction benchmarks.

Download Paper | Download Bibtex

sodaball

Publications

Conference Papers

Accurately Predicting Protein Mutational Effects via a Hierarchical Many-Body Attention Network

Quadruple Attention in Many-body Systems for Accurate Molecular Property Predictions