My research lies in the areas of large language models, feedback-driven generation, and the intersection of code generation and natural language reasoning. An overarching theme of my research is the use of inference-time compute for enhanced reasoning.