Aravindan Raghuveer

Algorithmic Guarantees for Distilling Supervised and Offline RL Datasets

Robust Reward Modeling via Causal Rubrics

Learning from Label Proportions: Bootstrapping Supervised Learners via Belief Propagation