Francis Bach

Fast kernel methods: Sobolev, physics-informed, and additive models

Scaling Laws for Gradient Descent and Sign Descent for Linear Bigram Models under Zipf’s Law

Backward Conformal Prediction

Convergence of the Gradient Flow for Shallow ReLU Networks on Weakly Interacting Data

Kernel Learning with Adversarial Features: Numerical Efficiency and Adaptive Regularization

Statistical Collusion by Collectives on Learning Platforms

The Surprising Agreement Between Convex Optimization Theory and Learning-Rate Scheduling for Large Model Training

Sampling Binary Data by Denoising through Score Functions