Hey, i'm Marmik

welcome to my web eden.

i work on mechanistic interpretability research, which i like to think of as the biology of artificial neural nets. i'm very interested in understanding how deep learning algorithms gives rise to its own world of beautiful structure and elegant complexity and behaviors.

recently, i've been working on understanding feature manifolds (how and why do they form) and interpreting genomic language models like Evo2.

in the past, i worked on research on interpreting Mixture of Experts (MoEs) for expert specialization under Arcadia Research Team and also collaborated with Insight + Interaction Lab at Harvard.

apart from work, i enjoy reading books of various genres (my bookshelf) and i've also been painting for a long time (my paintings). i've also been learning to play guitar, i like running and did a triathlon with my friends.

i like meeting new people and if anything from above is exciting to you, feel free to reach out to me on twitter or at marmik [at] berkeley [dot] edu