About
Hi, my name is Bishal Santra. I am a Senior Researcher at Microsoft Research India. My main area of research at MSRI lies at the intersection of information retrieval, synthetic data generation with LLMs, and sequential recommendation. I did my Ph.D. from IIT Kharagpur under the supervision of Prof. Pawan Goyal. My Ph.D. thesis is titled “From Dialog Representation to Response Generation: Deep Learning Techniques for Effective Dialog Modeling.”
I work on problems in retrieval and language modeling, and I’m particularly interested in how we can use LLMs to generate useful training data and build robust retrieval systems.
Research interests: ML for NLP, Information Retrieval.
Selected publications:
- SCULPT: Systematic Tuning of Long Prompts (ACL 2025).
- On the Necessity of World Knowledge for Mitigating Missing Labels in Extreme Classification (KDD 2025).
- Frugal Prompting for Dialog Models (EMNLP 2023).
- Representation Learning for Conversational Data using Discourse Mutual Information Maximization (NAACL 2022).