Hello -- I'm Dillon, a Computer Science PhD student at Duke University, advised by Ron Parr. Before graduate school, I worked as a business analyst and economic consultant.
I study Reinforcement Learning, focusing on the relationship between value function approximation, policy optimization, and the distribution of the training data.
My intended graduation year is 2028.
I study Reinforcement Learning, focusing on the relationship between value function approximation, policy optimization, and the distribution of the training data.
My intended graduation year is 2028.
Publications
- Approximate Next Policy Sampling: Replacing Conservative Target Policy Updates in Deep RL Dillon Sandhu, Ronald Parr preprint, forthcoming at RLC, 2026.
Blog
- Non-linear Value Learning — September 25, 2026
- Symmetrized TD Learning — July 07, 2026
- How reinforcement learning with TD learning depends on the inherent symmetry of the problem — June 09, 2026
- Will Deep TD Learning Result in Quality Representations? — May 20, 2026