Amir-massoud Farahmand

Relative Entropy Pathwise Policy Optimization

MAD-TD: Model-Augmented Data stabilizes High Update Ratio RL

Majority of the Bests: Improving Best-of-N via Bootstrapping

PANDAS: Improving Many-shot Jailbreaking via Positive Affirmation, Negative Demonstration, and Adaptive Sampling

A Truncated Newton Method for Optimal Transport

Calibrated Value-Aware Model Learning with Probabilistic Environment Models

Categorical Distributional Reinforcement Learning with Kullback-Leibler Divergence: Convergence and Asymptotics