I'm a recent McGill graduate (B.Sc. in Statistics and Computer Science) currently pursuing an MSc at Mila and Université de Montréal. I focus on AI safety research, particularly AI Control—developing protocols to safely deploy AI systems we don't fully trust. I also work at an AI safety consulting company.
Recent projects include: Untrusted Editing protocols & exploring feedback exploitation in control schemes, AI Control scaling laws, combining probe and black-box monitors for AI Control (NeurIPS poster), and currently, building AI Control evals.
My path into AI safety included the ML Safety Scholars program, leading an AI Safety Camp cyborgism project, building tools like AlignmentSearch and Stampy Chat to help others learn about alignment, and the MARS program. During my undergrad, I worked on sim-to-real transfer for PPO-based robotic locomotion under Prof. Hsiu-Chin Lin.
I organize ACX Montreal meetups and previously helped run EA McGill and found AI Alignment McGill. I
also enjoy chess, arm wrestling, and forecasting.
Projects
A real-time multiplayer infinite minesweeper game. Players explore an unbounded world together, competing on a global leaderboard.
Developed feedback-based AI control protocols
Open Low Cost Humanoid
Developing accessible, open-source humanoid robot with PPO-based locomotion for sim2sim and sim2real transfer
Lifelogging
High-performance Rust-based continuous audio recording with SIMD optimization
A retrieval-augmented generation platform that helps users explore AI safety research through conversational interface
I've also worked on some less serious projects: FriendBench (benchmarking AI friendliness), PressBench (benchmarking AI bench-press self-assessment), and an n-dimensional chess calculator.