← Home

Publications

EquiformerV3 architecture, equivariant graph attention, embedding, and feedforward network.

EquiformerV3: Scaling Efficient, Expressive, and General SE(3)-Equivariant Graph Attention Transformers

Yi-Lun Liao, Alexander J. Hoffman, Sabrina C. Shen, Alexandre Duval, Sam Walton Norwood, Tess Smidt

arXiv 2026 paper / code

ADiT overview: an autoencoder reconstructs molecules and materials, and a latent diffusion transformer generates new atomic structures.

All-atom Diffusion Transformers: Unified Generative Modelling of Molecules and Materials

Chaitanya K. Joshi, Xiang Fu, Yi-Lun Liao, Vahe Gharakhanyan, Benjamin Kurt Miller, Anuroop Sriram, Zachary W. Ulissi

ICML 2025 paper / code

DeNS overview: force encoding enables denoising non-equilibrium structures on a potential energy surface.

Generalizing Denoising to Non-Equilibrium Structures Improves Equivariant Force Fields

Yi-Lun Liao, Tess Smidt, Muhammed Shuaibi*, Abhishek Das*

TMLR 2024 (J2C Certification) paper / code

EquiformerV2 architecture and equivariant graph attention, highlighting separable layer normalization, eSCN convolution, and attention re-normalization.

EquiformerV2: Improved Equivariant Transformer for Scaling to Higher-Degree Representations

Yi-Lun Liao, Brandon Wood, Abhishek Das*, Tess Smidt*

ICLR 2024 paper / code

Figure 1: Equiformer architecture, equivariant graph attention, embedding, and feed-forward network.

Equiformer: Equivariant Graph Attention Transformer for 3D Atomistic Graphs

Yi-Lun Liao, Tess Smidt

ICLR 2023 (Spotlight) paper / code

Multi-stage Vision Transformer architecture with patch embeddings, transformer blocks, and residual spatial reduction.

Searching for Efficient Multi-Stage Vision Transformers

Yi-Lun Liao, Sertac Karaman, Vivienne Sze

arXiv 2021 paper / code

PARP overview comparing task-aware subnetwork discovery with prune, adjust, and re-prune.

PARP: Prune, Adjust and Re-Prune for Self-Supervised Speech Recognition

Cheng-I Jeff Lai, Yang Zhang*, Alexander H. Liu*, Shiyu Chang*, Yi-Lun Liao, Yung-Sung Chuang, Kaizhi Qian, Sameer Khurana, David Cox, James Glass

NeurIPS 2021 (Spotlight) paper

NetAdaptV2 overview: super-network training and multi-layer coordinate descent to discover a network.

NetAdaptV2: Efficient Neural Architecture Search with Fast Super-Network Training and Architecture Optimization

Tien-Ju Yang, Yi-Lun Liao, Vivienne Sze

CVPR 2021 paper