CS Research
Dynamic Programming, GPU + LLMs
My CS research spans two threads: how to teach dynamic programming (published in JCSC with Prof. Ali Erkan) and how LLMs behave on GPUs (ongoing with Prof. Venkata Govindarajan — the work behind my MLX contribution).
What I work on
With Prof. Ali Erkan I co-authored peer-reviewed work on dynamic programming pedagogy — whether top-down or bottom-up is the better way to teach it — published in the Journal of Computing Sciences in Colleges.
With Prof. Venkata Govindarajan I work on large language models and GPU programming: benchmarking, performance-oriented computing, and structured evaluation workflows. That performance work is what led to my merged optimization in Apple's MLX.