I am Abhinav, a Pre-Doctoral Research Fellow at Microsoft Research India , where I work with Dr. Amit Sharma, Dr. Nagarajan Natarajan, Dr. Srivathsan Koundinyan, and Prof. Vineeth N. Balasubramanian.

Over the past few years, I’ve been fortunate to learn from exceptional mentors and collaborators. At Microsoft Research India, I work on efficient grounding for retrieval-augmented models such as FrugalRAG and on building evaluation benchmarks for deep research like LiveDRBench. I also collaborate with Prof. Vineeth on understanding task transfer in vision–language models.

I received my B.Tech in Computer Science and Engineering from Delhi Technological University in 2022. During my undergraduate studies, I collaborated with the MIT Media Lab on privacy-preserving point cloud release.

I have been fortunate to be advised by Prof. Chirag Agarwal at the University of Virginia on methods for protecting textual data from unintended model training — our work on Unlearnable Text Datasets was accepted at NAACL 2025. Earlier, as a Research Associate at Adobe MDSR, I explored efficient multimodal learning for documents, exemplar-based image editing, and responsible AI with Balaji Krishnamurthy.

Broadly, I enjoy building practical methods that make machine learning systems more efficient, controllable, and trustworthy. I’m currently looking for Ph.D. opportunities; if you think there’s a good fit, I’d love to connect. Feel free to reach out via email.