University of Toronto
Computer Science & Mathematics
Education
2025 — 2029
Experience
Jun 2026 — Aug 2026
Algoverse
AI Research Intern
Research
-
A study of LLM hierarchies showing that competitive pressure can remain nearly invisible in an agent’s language while sharply changing how it allocates resources.
-
A controlled study of residual feedback and annealing schedules in adaptive state aggregation.
-
An empirical comparison of multithreaded value iteration and adaptive state aggregation on large tabular MDPs.
Projects
-
A two-time winning Hack Canada 2026 project.
- Winner — Backboard.io: Best Use of Backboard
- Winner — SPUR: Build a Real Canadian Startup (Top 3)
-
An experimental platform for studying how small populations of LLM agents can model and predict behaviour in social settings.
- Calculations & metrics (PDF)