Vashisth Tiwari

he/him ┃AI @CMU ┃Physics+Math @UofR

prof_pic.jpg

Hi! I’m Vashisth (or just Vash). I currently work as a Research Assistant with Prof. Emma Strubell where I am working on syntehtic data, mathematical reasoning, and green-Ai. In December 2024, I graduated from my master’s in the ECE Dept at Carnegie Mellon University, focusing on AI. Before this, I studied Physics and Math at the University of Rochester. I am broadly interested in improving LLMs in reasoning and data-scarce areas, model compression, and optimizing LLM inference.

I’ve had the privilege of working with exceptional mentors across physics, math, and AI. Currently, under the mentorship of Clara Na and Lucio Dery in Prof. Emma Strubell’s lab, I am exploring model-aware and active synthetic data generation. Previously, I worked with Prof. Beidi Chen on using speculative decoding to improve throughput and latency, self-speculation by utilizing weight, activation, and contextual sparsity.

My research journey began with my work in Dark Energy Spectroscopic Instrument with Prof. Segev BenZvi and with Prof. Machiel Blok on superconducting qubits.

Outside of academia, I’ve had the privilege of interning at Los Alamos National Laboratory under Dr. Malcolm Boshier where I worked on optimizing quantum systems. I’ve also gained valuable experience as a Quantitative Research Intern at Mana, mentored by Max Novendstern and David Kaufman.


/bits-&-bytes

Image 1 Image 2 Image 4 Image 3 Image 3 Image 3 Image 3 Image 3

When I'm not hunching over my laptop, you'll find me behind a camera lens—check out my photography portfolio. I spent much of my life playing badminton, including on the team at the University of Rochester, and now I'm attempting (and miserably failing) to learn tennis. You might also spot me tossing a frisbee at the CUT.

And yes, I love food and coffee, perhaps a bit too much. I'm always on the lookout for new spots, so if you have recommendations, send them my way!


news

Aug 19, 2024 ICLR’25 here we come MagicDec
Aug 19, 2024 Released MagicDec: Breaking Throughput-Latency Trade-off for Long Context Generation with Speculative Decoding. Blog
Aug 15, 2024 TAing Advanced NLP (11-711) with Prof. Graham Neubig
Jul 15, 2024 Started working in Prof. Emma Strubell’s Lab

latest posts