About me
Hi. I'm Uday Bhaskar.
I'm an RL Resident at Prime Intellect, working on automating post-training and reinforcement learning for language models. I think recursive self-improvement is the next frontier of scaling intelligence, and automating post-training is one of the key pieces needed to unlock RSI and autonomous continual learning. Prime Intellect is building open superintelligence infrastructure to democratize AI. I recommend trying their Lab.
AI is one of the most important technologies humanity will build, and it should benefit everyone. I care deeply about Intelligence Amplification of humans using AI. The problems I care about the most are recursive self-improvement of general intelligence and specialized AI systems for automating scientific discovery.
Previously, I worked on deep learning for high-risk applications at the Machine Learning Lab, IIIT Hyderabad. Before that, I pursued athletics for most of my teenage years and failed.
Links: Email, GitHub, HuggingFace, Google Scholar, X.
Credits: This website was inspired by Nat Friedman and Patrick Collison. Code written by ChatGPT.