Benjamin I. P. Rubinstein

Adaptive Data Analysis for Growing Data

AdaptDel: Adaptable Deletion Rate Randomized Smoothing for Certified Robustness

Multi-level Certified Defense Against Poisoning Attacks in Offline Reinforcement Learning

DDAD: A Two-Pronged Adversarial Defense Based on Distributional Discrepancy

One Stone, Two Birds: Enhancing Adversarial Defense Through the Lens of Distributional Discrepancy

TuBA: Cross-Lingual Transferability of Backdoor Attacks in LLMs with Instruction Tuning

Certified Adversarial Robustness via Randomized $\alpha$-Smoothing for Regression Models

Getting a-Round Guarantees: Floating-Point Attacks on Certified Robustness

The Certification Paradox: Certifications Admit Better Evasion Attacks