My name is Chris and I am a PhD-track MSc student at the University of British Columbia. My skills span networks, backend and applied artificial intelligence, all which I recently got to explore as a GenAI software engineering intern at Planview.

Last year, I contributed to and received an acknowledgement in the PLAICraft paper, a project collecting multimodal AI training data from Minecraft (see: PLAICraft project site). My undergraduate capstone project was like a fork of it called Project Elephant; it's a PLAICraft for everything. If it's a desktop app that can be accessed via a cloud machine, it can be part of a framework for scalable AI training data capture.

Currently, I conduct research on world action models (WAMs). I focus on positional embedding (RoPE variations) and latent diffusion optimizations. My immediate goal is to increase robustness in action-conditioned video frames tracking movement, such as ballistic movements of the mouse.