Himanshu Singhal

Research Fellow @Micrsoft Research

B.Tech Data Science & AI

Indian Institute of Technology Guwahati

himanshu.hs918@gmail.com

Hey! I’m Himanshu, a Research Fellow at Microsoft Research, working with the ReFoRM team. I graduated from the Indian Institute of Technology Guwahati with a degree in Data Science and AI, and I’m passionate about Machine Learning (ML), Deep Learning (DL), Natural Language Processing (NLP), Computer Vision (CV), and Large Language Models (LLMs).

My expertise lies in Agentic AI systems, Retrieval pipelines, and diffusion models. My Bachelor’s Thesis focused on learning richer representations for sequential/temporal data by reframing it through image-based representations, leveraging Vision Transformers. I’ve also worked on various projects, represented my college in two Inter IIT Tech Meets, and completed research internships at Arizona State University and Adobe Research.

My current research interests lie in Information Retrieval and developing effective representations for learning — building on my broader focus on efficient, scalable ML systems.

Connect with me @himanshu.hs918@gmail.com

news

Jul 12, 2026	Graduated from Indian Institute of Technology, Guwahati with a B.Tech. degree in Data Science and Artificial Intelligence.
Jul 06, 2026	Joined Microsoft Research as a Research Fellow working in the ReFoRM (Reimagining Frontier Large Retrieval Models) Team.
Apr 07, 2026	Paper titled *REaR: Retrieve, Expand and Refine for Effective Multitable Retrieval* accepted to ACL 2026 (Main Conference).
Oct 26, 2022	Started my Bachelors Degree at Indian Institute of Technology, Guwahati.

selected publications

ACL Main 2026
REaR: Retrieve, Expand and Refine for Effective Multitable Retrieval

Rishita Agarwal, Himanshu Singhal, Peter Baile Chen, and 3 more authors

ACL Main 2026, 2026

Abs Bib

Answering natural language queries over relational data often requires retrieving and reasoning over multiple tables, yet most retrievers optimize only for query-table relevance and ignore table–table compatibility. We introduce REaR (Retrieve, Expand and Refine), a three-stage, LLM-free framework that separates semantic relevance from structural joinability for efficient, high-fidelity multi-table retrieval. REaR (i) retrieves query-aligned tables, (ii) expands these with structurally joinable tables via fast, precomputed column-embedding comparisons, and (iii) refines them by pruning noisy or weakly related candidates. Empirically, REaR is retriever-agnostic and consistently improves dense/sparse retrievers on complex table QA datasets (BIRD, MMQA, and Spider) by improving both multi-table retrieval quality and downstream SQL execution. Despite being LLM-free, it delivers performance competitive with state-of-the-art LLM-augmented retrieval systems (e.g., ARM) while achieving much lower latency and cost. Ablations confirm complementary gains from expansion and refinement, underscoring REaR as a practical, scalable building block for table-based downstream tasks (e.g., Text-to-SQL).
@article{agarwal2025rear, title = {REaR: Retrieve, Expand and Refine for Effective Multitable Retrieval}, author = {Agarwal, Rishita and Singhal, Himanshu and Chen, Peter Baile and Choudhury, Manan Roy and Roth, Dan and Gupta, Vivek}, journal = {ACL Main 2026}, year = {2026}, url = {https://arxiv.org/abs/2511.00805} }

GAFSV-Net: A Vision Framework for Online Signature Verification

Himanshu Singhal and Suresh Sundaram

2026

Bib

@misc{singhal2026gafsvnetvisionframeworkonline,
  title = {GAFSV-Net: A Vision Framework for Online Signature Verification},
  author = {Singhal, Himanshu and Sundaram, Suresh},
  year = {2026},
  eprint = {2605.00120},
  archiveprefix = {arXiv},
  primaryclass = {cs.CV},
  url = {https://arxiv.org/abs/2605.00120}
}