Research Scientist, World Models

Waabi Waabi · Robotics · Toronto, ON +3 · Remote · Autonomy & Algorithms

Research Scientist role focused on developing and productionizing large-scale world models for temporal reasoning and generation in the context of autonomous transportation. This involves research in video generation, multimodal foundation models, LLM/VLM/VLA methods, and generative scenario modeling, with a strong emphasis on publishing research and integrating models into training pipelines.

What you'd actually do

  1. Conduct fundamental and applied research in generative and predictive world-modeling
  2. Collaborate with engineers to integrate models into large-scale, distributed training and rendering pipelines.
  3. Publish high-impact research at top conferences (CVPR, ECCV, ICCV, NeurIPS, ICLR, ICRA, SIGGRAPH).
  4. Mentor junior scientists and interns; foster a culture of scientific rigor and rapid experimentation.
  5. Stay on top of emerging advances in generative AI, differentiable rendering, knowledge distillation/compression, and robotics.

Skills

Required

  • Python
  • PyTorch
  • JAX
  • distributed training
  • generative models
  • predictive models
  • world models
  • temporal reasoning
  • video generation
  • multimodal foundation models
  • LLM
  • VLM
  • VLA
  • scene understanding
  • reasoning
  • control
  • generative scenario modeling
  • controllable simulation
  • model distillation

Nice to have

  • translate research into production-quality code
  • measurable product impact
  • first-author publications
  • differentiable rendering
  • knowledge distillation/compression
  • robotics

What the JD emphasized

  • Ph.D. in Computer Vision, Machine Learning, Robotics, or a related field or equivalent research experience pushing the boundaries of a technical field.
  • Expert domain knowledge: You have built generative or predictive models of the physical world with scale and efficiency in mind for real-world applications
  • Demonstrated technical innovation
  • Strong prototyping and implementation
  • Expert domain knowledge

Other signals

  • develop algorithms and productionize the next generation of World Models
  • large-scale world models for temporal reasoning and generation
  • video models, multimodal generative models, LLM/VLM/VLA models, and predictive models of traffic participants and scenes
  • publish high-impact research at top conferences