About

Chase Blagden is a ML research engineer at Aurora Flight Sciences with a B.S. in Information and Data Science from the California Institute of Technology (Caltech). Previously, he interned at Agentic AI, a startup working on automated video game testing, where he implemented offline RL techniques and an internal transformer implementation. Prior to that, he interned at Google, where he worked on calibrating models for the Hotel Ads team. Chase has also run his own business - he successfully bootstrapped a Solana RPC server rental business, where he wrote bots to handle user payments and usage as well as monitored server health.

Interesting papers:

• offline RL (including Q-transformer, ILQL, trajectory transformer, decision transformer)
• LLMs (Space and Time, General Pattern Machines, RLHF, QLoRa)
• robotics (RT-X, RT-2, diffusion policy, diffusion planner)