Yaodong Yu

Affiliations

EECS, UC Berkeley

papers

582

total citations

papers (13)

Learning Diverse and Discriminative Representations via the Principle of Maximal Coding Rate Reduction

NEURIPS 2020arXiv

234

citations

Token Statistics Transformer: Linear-Time Attention via Variational Rate Reduction

ICLR 2025arXiv

citations

ViP: A Differentially Private Foundation Model for Computer Vision

ICML 2024arXiv

citations

Masked Completion via Structured Diffusion with White-Box Transformers

ICLR 2024arXiv

citations

What You See is What You Get: Principled Deep Learning via Distributional Generalization

NEURIPS 2022arXiv

citations

A Global Geometric Analysis of Maximal Coding Rate Reduction

ICML 2024arXiv

citations

Differentially Private Representation Learning via Image Captioning

ICML 2024arXiv

citations

Adventurer: Optimizing Vision Mamba Architecture Designs for Efficiency

CVPR 2025arXiv

citations

Attention-Only Transformers via Unrolled Subspace Denoising

ICML 2025arXiv

citations

Yaodong Yu

Affiliations

papers (13)

Learning Diverse and Discriminative Representations via the Principle of Maximal Coding Rate Reduction

White-Box Transformers via Sparse Rate Reduction

Boundary thickness and robustness in learning models

Robust Calibration with Multi-domain Temperature Scaling

TCT: Convexifying Federated Learning using Bootstrapped Neural Tangent Kernels

Token Statistics Transformer: Linear-Time Attention via Variational Rate Reduction

ViP: A Differentially Private Foundation Model for Computer Vision

Masked Completion via Structured Diffusion with White-Box Transformers

What You See is What You Get: Principled Deep Learning via Distributional Generalization

A Global Geometric Analysis of Maximal Coding Rate Reduction

Differentially Private Representation Learning via Image Captioning

Adventurer: Optimizing Vision Mamba Architecture Designs for Efficiency

Attention-Only Transformers via Unrolled Subspace Denoising

papers (13)

Learning Diverse and Discriminative Representations via the Principle of Maximal Coding Rate Reduction

White-Box Transformers via Sparse Rate Reduction

Boundary thickness and robustness in learning models

Robust Calibration with Multi-domain Temperature Scaling

TCT: Convexifying Federated Learning using Bootstrapped Neural Tangent Kernels

Token Statistics Transformer: Linear-Time Attention via Variational Rate Reduction

ViP: A Differentially Private Foundation Model for Computer Vision

Masked Completion via Structured Diffusion with White-Box Transformers

What You See is What You Get: Principled Deep Learning via Distributional Generalization

A Global Geometric Analysis of Maximal Coding Rate Reduction

Differentially Private Representation Learning via Image Captioning

Adventurer: Optimizing Vision Mamba Architecture Designs for Efficiency

Attention-Only Transformers via Unrolled Subspace Denoising