eigenron

i'm glad you're here :)

i work on scaling RL, advancing reasoning gaming agents, transfer learning, and world models.

i studied physics, math and computer science (and often, philosophy) in my undergrad. i've researched phase qubits for my senior (theoretical) research, and currently working at a startup, on ml infra!

i love ablation and coding things up from scratch. things i've built: GPT-2, Llama-2/3, Stable Diffusion-v1.5, Diffusion-based ESM, Physics-aware Autograd Engine, VAE + MDN-RNN World Model, Chess Engine, etc.

i like to tweet: @eigenron

Note: my site was previously this, but i'm moving it here, and thus this site is a work in progress :)