Nikita Balagansky
I did a Ph.D. at MIPT. I’m interested in general representation learning, especially for LLMs, including mechanistic interpretability.
NEW Here is an interactive blogpost on counting manifolds made by Slava and me.
I did a Ph.D. at MIPT. I’m interested in general representation learning, especially for LLMs, including mechanistic interpretability.
NEW Here is an interactive blogpost on counting manifolds made by Slava and me.