Jefferson R

Focused on RLHF and human preference data for Portuguese (PT-BR). Building datasets for training reward models and improving AI communication quality. Interested in DPO, NLP, and multilingual LLMs.