research interests

Current Research Interests

• Discrete and Continuous Flow Matching / Diffusion Models: I study both discrete-time and continuous-time formulations of flow matching and diffusion models, with a focus on probability path design, transport dynamics, and scalable training/sampling algorithms for high-dimensional or non-Euclidean data.
• Reinforcement Learning for Generative Models: I am interested in using RL methods to improve generative modeling pipelines, including reward-driven fine-tuning, policy optimization for generation quality/control, and learning strategies that connect exploration with robust sample synthesis.
• Reinforcement Learning for Decision-Making Over Time (toward quant research): I work on RL for sequential and temporal decision problems under uncertainty, with long-term interests in applications aligned with quantitative researcher roles in hedge funds and related data-driven decision-making jobs.