Yejin Choi

GPT-5-Pro proved a better bound on a convex optimization open problem than the published paper, improving the stepsize condition from 1.75/L to 1.5/L with a novel and verified proof.

“Her TED Talk is referenced as insightful commentary on GPT-4's capabilities and limitations in common sense reasoning”

TTT-Discover enables large language models to perform reinforcement learning at test time, adapting to problem-specific tasks and achieving state-of-the-art results in mathematics, GPU kernels, algorithms, and biology.

“Co-authored TTT-Discover reinforcement learning system for test-time LLM training.”