Yuyang Hu

RL, small parameter models, model optimization, diffusion (applying MOE to diffusion models), fine tuning large models to mimic individuals, building agents and pairing them up with RAG systems