Qasim Ali

prof_pic.jpg

Hi! I’m Qasim Ali, a graduate student at the University of Waterloo studying AI. I am advised by Alex Wong and Javad Shafiee in the Systems Engineering Department.

I’m interested in building and using World Models 🌎 to help agents take better actions. While LLMs are great at solving simple tasks, I think System 2 Reasoning 🧠 will require alternative architectures like JEPA or Neural Turing Machines. During my master’s degree, I worked at the intersection of generative models, reinforcement learning, and representation learning on ideas related to this vision.

I come from an embedded software and robotics background. I did my undergrad in Mechatronics Engineering from McMaster University and interned at Magna Electronic’s ADAS division. During that time I worked with various embedded software platforms to help get self-driving features working real-time on vehicles, built a control algorithm for getting vehicles to navigate through V2I-connected intersections, and built a drone.

If you’re into AI research, robotics, or just want to chat about the future of agents 🤖 — feel free to connect.

P.S. I’m graduating soon — if you’re looking for people in AI who can train, finetune or deploy models, then feel free to contact me.

news

Aug 08, 2025 I will be at RLC 2025 Workshop on Reward-free Reinforcement Learning, presenting MR-CRL! Hope to see you all in Edmonton!
Jul 18, 2025 I will be at ICML 2025 Workshop on Building Physicall Plausible World Models, presenting Humanoid World Models! Hope to see you all in Vancouver!
Jun 15, 2025 I will be at CVPR 2025 Workshop on 3D-LLM/VLA, presenting GraphPad. Hope to see you all in Nashville!
May 01, 2025 I finished second in the 1X World Model Challenge.

selected publications

  1. Humanoid World Models: Open World Foundation Models for Humanoid Robotics
    Muhammad Qasim Ali, Aditya Sridhar, Shahbuland Matiana, and 2 more authors
    2025
  2. GraphPad: Inference-Time 3D Scene Graph Updates for Embodied Question Answering
    Muhammad Qasim Ali, Saeejith Nair, Alexander Wong, and 2 more authors
    2025