I am a second year PhD student in Computer Science and Engineering at UC San Diego, where I am fortunate to be advised by Prof. Earlence Fernandes.
I work in AI security. Specifically, I am interested in approaches to create efficient attacks on LLMs and building principled defences for LLM-integrated systems.
I got my first experience doing security research working with Prof. Vinod Ganapathy where we built a system to enforce downstream control for distributed applications running on ROS2.
I received my BS and MS in mathematics from the Indian Institute of Science in Bangalore - working mostly on topology and geometry.
I also had a brief stint in industry at Datapoem building AI to address the pain points of the marketing industry.
Publications and Preprints
-
Fun-tuning: Characterizing the Vulnerability of Proprietary LLMs to Optimization-based Prompt Injection Attacks via the Fine-Tuning Interface
Andrey Labunets, Nishit V. Pandya, Ashish Hooda, Xiaohan Fu, Earlence Fernandes.
IEEE Symposium on Security and Privacy (S&P) 2025 -
Decentralized Information Flow Control for ROS2
Nishit V. Pandya, Himanshu Kumar, Gokulnath M. Pillai, Vinod Ganapathy.
Network and Distributed System Security Symposium (NDSS) 2024