Ayush Sekhari

Computationally Efficient RL under Linear Bellman Completeness for Deterministic Dynamics

The Gaussian Mixing Mechanism: Renyi Differential Privacy via Gaussian Sketches

Machine Unlearning Fails to Remove Data Poisoning Attacks

GaussMark: A Practical Approach for Structural Watermarking of Language Models

System Aware Unlearning Algorithms: Use Lesser, Forget Faster

System-Aware Unlearning Algorithms: Use Lesser, Forget Faster

Unstable Unlearning: The Hidden Risk of Concept Resurgence in Diffusion Models

Offline Data Enhanced On-Policy Policy Gradient with Provable Guarantees

Harnessing Density Ratios for Online Reinforcement Learning