Michael A Osborne

A Physics-Inspired Optimizer: Velocity Regularized Adam

SOReL and TOReL: Two Methods for Fully Offline Reinforcement Learning

Leave one Expert Out: Robust Uncertainty Quantification via Intrinsic Cross-Validation

Scalable Valuation of Human Feedback through Provably Robust Model Alignment

Just One Layer Norm Guarantees Stable Extrapolation

Q-based Variational Inverse Reinforcement Learning

Social Bayesian Optimization for Building Truthful Consensus