Nitish Shukla

Ph.D. student at Michigan State University in iPRoBe lab with Prof. Arun Ross.

My current research focuses on facial demorphing and preference optimization in multimodal large language models (MLLMs). In particular, my current research addresses two core questions: (a) how to optimize MLLMs for multi-image contexts without relying on extensive human annotations, and (b) how to leverage MLLMs to learn richer representations that improve performance on downstream retrieval tasks.

Before that, I studied M.Sc. Computer Science at Chennai Mathematical Institute and M.Sc. Mathematics at IIT Guwahati. At CMI, I worked with Prof. KV and Madhavan Mukund studying the robustness properties of DNNs induced by prototypes.

Outside academia, I find joy in cooking (My Instagram), watching movies, and playing video games during my free time. Beyond that, I’m an enthusiastic hiker and hold a deep admiration for snowy mountains. If there’s anything specific you’d like to discuss or share, feel free to contact me!