I am a researcher studying and designing deep neural networks. My work focuses on developing efficient machine learning algorithms
that run practically on existing hardware, guided by mathematical and physical principles.
Currently, I explore adaptive compute for transformers, inference-time reasoning, and world models.
Since fall 2024, I am pursuing a PhD under François Fleuret in
Geneva. Before that, I received an education in maths, physics, and computer science in prestigious institutions in France and the US.
My work involves programming, scribbling maths and diagrams, and befriending the LLMs.
Besides this exciting main quest, I am a ski instructor in val d'isère and a dj (currently on sabattical) playing underground electronic music.
I also enjoy nature, spending time with friends, and traveling alone.
Finally, I believe in open knowledge, ai decentralization, and the fall of attention merchants.
I am open to collaborating, mentoring, and being mentored. Reach out.