Hi - I'm Lucas and I'm from Nebraska.
I'll spare you the details but here's my quick gist:
Currently:
At UCSD studying ML and doing research applying RL to LLMs. My current work is focused on building a corporate sim to use as a multi-turn RL environment for LLMs. Recently released some work I led (+ gave a talk @ NeurIPS) on how your fine-tuning data influences downstream reasoning in chessI promise I'm more of a person than this - feel free to reach out! 👋