Picture for Zhenyu Liao

Zhenyu Liao

Random Matrix Theory for Deep Learning: Beyond Eigenvalues of Linear Models

Add code
Jun 16, 2025
Viaarxiv icon

Fundamental Bias in Inverting Random Sampling Matrices with Application to Sub-sampled Newton

Add code
Feb 19, 2025
Viaarxiv icon

A Large-dimensional Analysis of ESPRIT DoA Estimation: Inconsistency and a Correction via RMT

Add code
Jan 06, 2025
Viaarxiv icon

The Breakdown of Gaussian Universality in Classification of High-dimensional Mixtures

Add code
Oct 08, 2024
Viaarxiv icon

"Lossless" Compression of Deep Neural Networks: A High-dimensional Neural Tangent Kernel Approach

Add code
Mar 01, 2024
Viaarxiv icon

Deep Equilibrium Models are Almost Equivalent to Not-so-deep Explicit Models for High-dimensional Gaussian Mixtures

Add code
Feb 05, 2024
Viaarxiv icon

Robust and Communication-Efficient Federated Domain Adaptation via Random Features

Add code
Nov 08, 2023
Viaarxiv icon

On the Equivalence between Implicit and Explicit Neural Networks: A High-dimensional Viewpoint

Add code
Aug 31, 2023
Viaarxiv icon

Analysis and Approximate Inference of Large and Dense Random Kronecker Graphs

Add code
Jun 14, 2023
Viaarxiv icon

Semantic Image Manipulation with Background-guided Internal Learning

Add code
Mar 24, 2022
Figure 1 for Semantic Image Manipulation with Background-guided Internal Learning
Figure 2 for Semantic Image Manipulation with Background-guided Internal Learning
Figure 3 for Semantic Image Manipulation with Background-guided Internal Learning
Figure 4 for Semantic Image Manipulation with Background-guided Internal Learning
Viaarxiv icon