Song, Zhao
137 publications
AISTATS
2025
Bypassing the Exponential Dependency: Looped Transformers Efficiently Learn In-Context by Multi-Step Gradient Descent
NeurIPS
2025
Differential Privacy for Euclidean Jordan Algebra with Applications to Private Symmetric Cone Programming
AISTATS
2025
Fourier Circuits in Neural Networks and Transformers: A Case Study of Modular Arithmetic with Multiple Inputs
ICML
2025
Fundamental Limits of Visual Autoregressive Transformers: Universal Approximation Abilities
ICLRW
2025
How Sparse Attention Approximates Exact Attention?Your Attention Is Naturally $n^C$-Sparse
CPAL
2025
The Computational Limits of State-Space Models and Mamba via the Lens of Circuit Complexity
ICLRW
2025
Video Latent Flow Matching: Optimal Polynomial Projections for Video Interpolation and Extrapolation
AISTATS
2025
When Can We Solve the Weighted Low Rank Approximation Problem in Truly Subquadratic Time?
NeurIPS
2024
Log-Concave Sampling from a Convex Body with a Barrier: A Robust and Unified Dikin Walk
NeurIPS
2024
On Statistical Rates and Provably Efficient Criteria of Latent Diffusion Transformers (DiTs)
AISTATS
2023
A Tale of Two Efficient Value Iteration Algorithms for Solving Linear MDPs with Large Action Space
NeurIPS
2023
Bypass Exponential Time Preprocessing: Fast Neural Network Training via Weight-Data Correlation Preprocessing
NeurIPS
2023
InfoPrompt: Information-Theoretic Soft Prompt Tuning for Natural Language Understanding
ICML
2023
Sketching Meets Differential Privacy: Fast Algorithm for Dynamic Kronecker Projection Maintenance
ICML
2023
Sketching for First Order Method: Efficient Algorithm for Low-Bandwidth Channel and Vulnerability
UAI
2021
When Is Particle Filtering Efficient for Planning in Partially Observed Linear Dynamical Systems?