Deep learning research student focused on representation learning, self-supervised learning, and world models.
Currently seeking a 6-month final-year research internship in deep learning starting in 2027.
- Opportunistic Target Selection — CAp 2026 accepted paper on query-efficient black-box adversarial attacks.
- SLS-WM — Ongoing research on structured label smoothing for discrete world models.
- Hack the World(s) — Finalist at a Yann LeCun–sponsored JEPA/world-model hackathon; built a self-supervised EEG system with Riemannian analysis of the learned representations.
- Rose — Open-source League of Legends customization tool with 15K+ daily active users.