Independent ML researcher · Sofia, Bulgaria
Vasil
Vasilev.
Researching how smaller models learn useful representations from less supervision.
01 / About
Curious about how machines learn.
My interests span self-supervised learning, efficient AI, and representation learning for speech, vision, and world models. I enjoy working across research and implementation, from experiments to the systems that make them possible.
This page is a starting point. More about my research and interests is on the way.
02 / Selected work
Projects & research.
- Local Repair in Medical VLMs Code and reproducibility artifacts for a controlled study of downstream reliability.
- GeoJEPA Explore the project on GitHub.
- Medical-VLM Explore the project on GitHub.
Publication details will be added once the correct Scholar profile is confirmed.
03 / Elsewhere