Chenlu Ye

Transformers as Multi-task Learners: Decoupling Features in Hidden Markov Models

Daunce: Data Attribution through Uncertainty Estimation

Beyond Correctness: Harmonizing Process and Outcome Rewards through RL Training

Logarithmic Regret for Online KL-Regularized Reinforcement Learning

Sharp Analysis for KL-Regularized Contextual Bandits and RLHF

Catoni Contextual Bandits are Robust to Heavy-tailed Rewards

Understanding Overadaptation in Supervised Fine-Tuning: The Role of Ensemble Methods

Sharp Analysis for KL-Regularized Contextual Bandits and RLHF