Research

Authors indicated by a (*) indicate equal contribution listed in alphabetical order, as is customary in theoretical computer science.

Preprints

Conference and journal publications

2025

  1. Weak Poincaré Inequalities, Simulated Annealing, and Sampling from Spherical Spin Glasses
    ACM Symposium on Theory of Computing, 2025
  2. ICLR
    Provable Weak-to-Strong Generalization via Benign Overfitting
    David X Wu, and Anant Sahai
    International Conference on Learning Representations (Preliminary version at NeurIPS’24 M3L Workshop), 2025

2024

  1. Locally Stationary Distributions: A Framework for Analyzing Slow-Mixing Markov Chains
    IEEE Annual Symposium on Foundations of Computer Science, 2024
  2. Fast Mixing in Sparse Random Ising Models
    Kuikui Liu*Sidhanth Mohanty*Amit Rajaraman*, and David X Wu*
    IEEE Annual Symposium on Foundations of Computer Science, 2024
  3. Robust recovery for stochastic block models, simplified and generalized
    Sidhanth Mohanty*Prasad Raghavendra*, and David X Wu*
    ACM Symposium on Theory of Computing, 2024

2023

  1. Precise Asymptotic Generalization for Multiclass Classification with Overparameterized Linear Models
    David X Wu*, and Anant Sahai
    Neural Information Processing Systems, 2023
  2. On the Training Instability of Shuffling SGD with Batch Normalization
    David X Wu*Chulhee Yun, and Suvrit Sra
    International Conference on Machine Learning, 2023
  3. Lower Bounds for Multiclass Classification with Overparameterized Linear Models
    David X Wu*, and Anant Sahai
    International Symposium on Information Theory, 2023

2022

  1. Maximum A Posteriori Inference of Random Dot Product Graphs via Conic Programming
    David X Wu*David Palmer, and Daryl R DeFord
    SIAM Journal on Optimization, 2022