Explore projects

BIGE is a framework for generative models to adhere to clinician-defined constraints. To generate realistic motion, our method uses a biomechanically informed surrogate model to guide the generation process.

official repository for “Reinforcement Learning for Reasoning in Large Language Models with One Training Example”

Are models trained to use test time chains of thought before answering any safer?

Long Lived Particle Ntupler based on AOD