Shobhit Agarwal

Hi, I'm Shobhit! I'm a student at Stanford studying Computer Science, Math, and Philosophy. I work on interpretability, reinforcement learning, computer vision, and applications of AI in healthcare.

Currently, I'm part of the engineering team @ Ambience Healthcare (one of the first hires: trained early clinical AI models), researching video-language models @ Stanford Vision & Learning Lab, post-training specialized LLMs for radiology report generation @ Harvard, and TA for Stanford's Intro CS Classes.

Previously, I helped organize and build core infrastructure for TreeHacks, the largest collegiate hackathon in the world, architected internal benchmarking tools for BeHeard Labs, and studied quantitative finance @ Jane Street.

Always looking for new people to chat with! Feel free to reach out anytime :)

shobhit [dot] agarwal [at] stanford [dot] edu
github / linkedin / google scholar

Research

SpliceFT: Splicing for Efficient Video Model Training
Zane Durante, Silky Singh, Arpandeep Khatua, Shobhit Agarwal, Reuben Tan, Yong Jae Lee, Jianfeng Gao, Ehsan Adeli, Fei-Fei Li
[REDACTED], 2026
Representing visual classification as a linear combination of words
Shobhit Agarwal, Yevgeniy R. Semenov, William Lotter
Machine Learning for Healthcare Conference (ML4H), 2023
Self-supervised deep learning to predict molecular markers from routine histopathology slides for high-grade glioma tumors
Olivia Krebs, Shobhit Agarwal, Pallavi Tiwari
SPIE Medical Imaging, 2023
A multimodal machine learning approach to diagnosis, prognosis, and treatment prediction for neurodegenerative diseases and cancer
Shobhit Agarwal
IEEE UEMCON, 2022

Notable Awards & Honors

Kleiner Perkins Fellowship
2025
ACM Cutler Bell Prize
2024
Coca-Cola Scholar
2024
International Science & Engineering Fair 2nd Grand Prize Winner
2024