Scholar
Prateek Verma
Google Scholar ID: Ta5-aX4AAAAJ
IIT Bombay, Stanford, Google [X]
Citations & Impact
Citations
592
H-index
12
i10-index
13
Publications
20
Co-authors
0
Contact
No contact links provided.
Publications
Browse publications on Google Scholar (top-right) ↗
Resume (English only)
Academic Achievements
- Upcoming paper; ICASSP submission on switching learned machine ears every 25ms based on input waveform content; some LLM and Transformer research featured on Stanford's YouTube channel with over 5,100 views; supported by generous Google cloud computing grants from Stanford Human-Centered AI Institute.
Research Experience
- Researcher at Stanford University; AI Resident at Google X, working on undisclosed projects; Acoustic Engineer at Stanford; Resident at Google X, Alphabet Inc.
Education
- Indian Institute of Technology (IIT) Bombay, M. Tech in Electrical Engineering with a specialization in Signal Processing.
Background
- Research interests include: advancing the creativity of computers in audio, text, music, and video generation; understanding and advancing our perception of the world through modern advancements; conducting classic signal processing research; advancing large language models across modalities; using inductive biases to make pretraining up to 50% faster; incorporating really long context >1M tokens; utilizing the internal circuitry of these models. His recent goal is to make computers hear with the best possible ears surpassing what evolution gave us.
Miscellany
- Likes running and biking; actively played field hockey, cricket, and table-tennis during IIT Bombay days; enjoys all kinds of music and is particularly impressed by the compositions of Rohail Hyatt and Trent Reznor-Atticus Ross.
Co-authors