Valentin Six
Hi there ! 👋🏼
My name is Valentin and I do research in Machine Learning.
My current research interests lie in the foundations and interpretability of generative models: understanding their inner mechanisms, both during training and inference.
I am currently a research intern in the Empirical Inference Lab at the Max Planck Institute for Intelligent Systems. I work with Prof. Bernhard Schölkopf and Lancelot Da Costa on using LLMs to help agents learn and plan effectively in the world they interact with.
In Fall 2026, I will begin a PhD in Computer Science at the University of Virginia, where I will be advised by Prof. Hadi Daneshmand. I was very fortunate to receive the UVA Provost Fellowship for my PhD.
Previously, I earned a Master’s degree in Computer Science from Georgia Tech, as well as Bachelor’s degree from Télécom SudParis, where I studied Mathematics and Physics.
Feel free to reach out!
latest posts
| Jul 07, 2026 | In-Context Learning: Learning at Inference Time |
|---|---|
| Jul 03, 2026 | A Gentle Introduction to Mechanistic Interpretability |
| Jul 02, 2026 | Singular Learning Theory and Developmental Interpretability |