Work

Research

My research focuses on learning and reasoning in large output spaces, mechanistic interpretability, and model distillation — advancing how models attend, reason, and generalize.