Sudhanshu Agrawal

M.S. Computer Science @ Stanford

Hi everyone! My name is Sudhanshu and I'm an incoming Master's in Computer Science student at Stanford. My current research interests are in studying the internal representations of large generative models, multimodality, and human-computer interaction.

Previously, I was an ML Research Engineer at Qualcomm AI, working on LLM efficiency research supervised by Mingu Lee. At Qualcomm, we worked on inventing new speculative decoding algorithms for edge applications — making LLMs fast enough to run locally on your phone or laptop!

I graduated from UCLA in 2023 with a double major in Computer Science and Mathematics. While I was at UCLA, I was fortunate to conduct research with Professor Aditya Grover on generative modeling and with Professor Levon Nurbekyan and Professor Samy Wu Fung on mean-field games.

In my free time, I like to surf, play the guitar, and sing. I also love watching movies, reading, and listening to music. Feel free to reach out if you'd like to chat!

Experience

ML Research Engineer

Qualcomm AI Research
LLM efficiency, speculative decoding, efficient architectures, diffusion LLMs.
2023 - Present
Qualcomm

ML Engineering Intern

Qualcomm AI Research
Profiling tools for deep learning applications.
Summer 2022
Qualcomm

ML Engineering Intern

SonicJobs
Synthetic computer vision dataset creation.
Summer 2021
SonicJobs

ML and Data Science Intern

Reliance Jio
Hydrocarbon property prediction using classical ML.
Summer 2020
Reliance Jio

ML Intern

Julia Computing Inc
Contributions to the Flux Model Zoo library.
Summer 2019
Julia Computing

Education

Bachelor of Science, Computer Science

University of California, Los Angeles (UCLA)
2019 - 2023
Magna cum laude
UCLA

Bachelor of Science, Mathematics

University of California, Los Angeles (UCLA)
2019 - 2023
Cum laude
UCLA

ISC 12th Grade

Mallya Aditi International School, Bengaluru
2017 - 2019
National Rank 4
Mallya Aditi International School

Publications

Google Scholar:
ICLR Sci4DL Workshop 2026

A Comparative analysis of Layer-wise Representational Capacity in AR and Diffusion LLMs

Comparing layer-wise representations in autoregressive vs. diffusion LLMs.
Raghavv Goel, Risheek Garrepalli, Sudhanshu Agrawal, Chris Lott, Mingu Lee, Fatih Porikli
2026
Layer-wise Representational Capacity Publication
ICML FoGen and SPIGM Workshops 2026

Structuring The Future: Diffusion LLM Speculative Decoding via Calibrated Draft Graphs

Novel speculative decoding algorithm to accelerate diffusion LLMs.
Sudhanshu Agrawal, Risheek Garrepalli, Raghavv Goel, Christopher Lott, Fatih Porikli, Mingu Lee
2025
Spiffy Publication
ICML ES-FoMo Workshop, 2025

VOCABTRIM: Vocabulary Pruning for Efficient Speculative Decoding in LLMs.

Reducing the vocabulary size of the draft model to reduce memory-bandwidth overhead during speculative decoding.
Raghavv Goel, Sudhanshu Agrawal, et al.
2025
VOCABTRIM Publication
NeurIPS ENLSP-IV Workshop, 2024

AdaEDL: Early Draft Stopping for Speculative Decoding of Large Language Models via an Entropy-based Lower Bound on Token Acceptance Probability

Early-draft-stopping using entropy for efficient speculative decoding.
Sudhanshu Agrawal, Wonseok Jeon, Mingu Lee
2024
AdaEDL Publication
NeurIPS, 2023

ExPT: Synthetic Pretraining for Few-Shot Experimental Design

Foundation model architecture for in-context adaptation to experimental design objectives.
Tung Nguyen, Sudhanshu Agrawal, Aditya Grover
2023
ExPT Publication
Journal of Computational Physics, 2022

Random Features for High-Dimensional Nonlocal Mean-Field Games

Using random-feature kernels to model mean-field interactions efficiently high-dimensional settings.
Sudhanshu Agrawal*, Wonjun Lee*, Samy Wu Fung, Levon Nurbekyan
2022
Mean-Field Games Publication

Patents

The following patent applications relate to 11 distinct inventions. Each invention has multiple US and global pending patent applications.

  1. TW 115121889(lead inventor)
  2. WO PCT/US2026/030830(lead inventor)
  3. WO PCT/US2026/030887(lead inventor)
  4. US 64/074,098(lead inventor)
  5. US 64/072,715(lead inventor)
  6. US 19/662,922
  7. US 19/634,806
  8. US 19/571,826(lead inventor)
  9. US 19/554,427(lead inventor)
  10. TW 115104375
  11. WO PCT/US2026/013469
  12. US 63/964,912
  13. US 63/944,624 (lead inventor)
  14. US 63/938,500 (lead inventor)
  15. WO PCT/CN2025/134909
  16. WO PCT/CN2025/124672
  17. US 63/872,751
  18. US 63/849,613 (lead inventor)
  19. US 19/273,664 (lead inventor)
  20. WO PCT/US2025/037170 (lead inventor)
  21. US 19/086,578 (lead inventor)
  22. US 18/983,103 (lead inventor)
  23. US 63/688,654 (lead inventor)

Blog

Medium

Generative AI for Experimental Design

Using generative modeling to solve offline black-box optimization problems.
2024
ExPT Blog
Medium

100-Dimensional Games

Understanding and solving nonlocal mean-field games
2023
Mean-Field Games Blog
FluxML.ai

Simulating The Motion of Charged Bodies

Simulating an N-body problem using gradient descent.
2023
Mean-Field Games Blog

Invited Talks | Judgeships | Reviewing

  • Reviewer: 2026 ICML Workshop on the Foundations of Generative Models
  • Reviewer: 2026 Transactions on Pattern Analysis and Machine Intelligence
  • Reviewer: 2026 AAAI Conference
  • Judge: 2025 UCSD Graduate Student Research Exposition
  • Judge: 2025 San Diego State University Student Research Symposium
  • Reviewer: 2025 ICML Efficient Systems for Foundation Models Workshop
  • Reviewer: 2025 NeurIPS Efficient Natural Language and Speech Processing Workshop
  • Speaker: 2024 UCSD and Qualcomm Graduate Students Tech Talk and Recruitment Event
  • Speaker: 2024 UCSD, IEEE, Qualcomm Careers Panel
  • Speaker: 2024 UCLA Mathematics Department Alumni Panel