Michael Muehlebach

The Sample Complexity of Online Reinforcement Learning: A Multi-model Perspective

Efficient Diffusion Models under Nonconvex Equality and Inequality constraints via Landing

ESLM: Risk-Averse Selective Language Modeling with Hierarchical Batch Selection

Fast Non-Log-Concave Sampling under Nonconvex Equality and Inequality Constraints with Landing

Quantization-Free Autoregressive Action Transformer

Distributed Event-Based Learning via ADMM

Zeroth-Order Optimization Finds Flat Minima

Partially Observable Reinforcement Learning with Memory Traces

Conformal Generative Modeling with Improved Sample Efficiency through Sequential Greedy Filtering

Adversarial Training for Defense Against Label Poisoning Attacks

A Pontryagin Perspective on Reinforcement Learning