about_me
Hi! I'm Sumit! I'm a machine learning engineer (day job), and an independent machine learning researcher focusing on efficient language models & low-resource language research.
I have a strong desire to contribute to the field of ML research & development to make the world a better place. You can learn more about my research interests in my research site if you're interested. But regardless, welcome!
You can learn more about me here where I go into further detail about my work and interests. Only if you're interested, of course.
If you'd like to get in touch, you can find me on LinkedIn or send me an email. I love talking over coffee, so I'm always up for a cup and chat if you're around the Tokyo area.
current_focus
| Studying | ML kernel optimization, HPC, and more | ongoing |
| Researching | Low-resource languages | in-progress |
| Learning by doing | Weekly hackathons and events in Tokyo (join me!) | active |
featured_blogs
featured_projects
LoRA and Friends
A controlled comparison of attention-only and all-layer LoRA on Qwen3-8B using an OpenMathInstruct-2-derived math SFT dataset and GSM8K evaluation
May 17, 2026
Expert Emergence in a Small Sparse MoE Transformer
Emergence of expertise across different domains when replacing a dense FFN with a MoE
Feb 26, 2026