Harsh Chaudhari
PhD Student
Biography
Harsh Chaudhari is a computer science doctoral student at Northeastern University, where he is advised by Alina Oprea. Chaudhari's research interests broadly lie at the intersection of machine learning, security, and privacy. His current research focuses on designing attacks on machine learning models to understand and mitigate threats to participating users.
Recent publications
-
Exploiting Leaderboards for Large-Scale Distribution of Malicious Models
Citation: Anshuman Suri, Harsh Chaudhari, Yuefeng Peng, Ali Naseh, Alina Oprea, Amir Houmansadr. (2026). Exploiting Leaderboards for Large-Scale Distribution of Malicious Models SP, 175-194. https://doi.org/10.1109/SP63933.2026.00044 -
Cascading Adversarial Bias from Injection to Distillation in Language Models
Citation: Harsh Chaudhari, Jamie Hayes, Matthew Jagielski, Ilia Shumailov, Milad Nasr, Alina Oprea. (2025). Cascading Adversarial Bias from Injection to Distillation in Language Models CCS, 4409-4422. https://doi.org/10.1145/3719027.3765122 -
Riddle Me This! Stealthy Membership Inference for Retrieval-Augmented Generation
Citation: Ali Naseh, Yuefeng Peng, Anshuman Suri, Harsh Chaudhari, Alina Oprea, Amir Houmansadr. (2025). Riddle Me This! Stealthy Membership Inference for Retrieval-Augmented Generation CCS, 1245-1259. https://doi.org/10.1145/3719027.3744840 -
Chameleon: Increasing Label-Only Membership Leakage with Adaptive Poisoning
Citation: Harsh Chaudhari, Giorgio Severi, Alina Oprea, Jonathan R. Ullman. (2024). Chameleon: Increasing Label-Only Membership Leakage with Adaptive Poisoning ICLR. https://openreview.net/forum?id=4DoSULcfG6
