Neelesh Bisht
ML Researcher. Engineer.
I’m a Master’s student at Carnegie Mellon University, where I do research in ML interpretability at the School of Computer Science. My work focuses on understanding what neural networks learn — building methods that surface meaning in model representations.
Before CMU, I spent three years at AiDash as an early engineer, building distributed data systems and infrastructure for large-scale geospatial workloads, as well as their geospatial web product.
I like working at the intersection of research and systems: understanding models deeply, and building the infrastructure that makes that work possible.
Selected Work(s)
CCDiff: Inverse-CCA for Discovering Visual Differences in Natural Language [Link]
Neelesh Bisht, X. Li, M.R. Uddin, Z. Li, Y. Liu, M. Xu
Accepted at ICML EMMQA Workshop 2026; Submitted to NeurIPS 2026
DiffCAM: Data-Driven Saliency Maps by Capturing Feature Differences
X. Li, Q. Zhao, Neelesh Bisht, M.R. Uddin, J.Y. Kim, B. Zhang, M. Xu
CVPR 2025 (Highlights, Top 2.6%)