Research Scientist, develop predictive world models from video using Python and PyTorch within Meta’s Fundamental AI Research team. Build productive algorithms for planning and control. This Research Scientist role advances artificial intelligence through world modelling and embodied agents. You will research self-supervised video learning, predictive models, model-based reinforcement learning, and model-predictive control. The work spans data curation, large-scale model training, and robust benchmarking. Responsabilités principales:

Lead research advancing world modelling and artificial intelligence Develop architectures and training methods for predictive environments Explore planning, prediction, control,



and decision-making applications Publish research at peer-reviewed venues Collaborate with scientists and engineers across teams Exigences:

Bachelor’s degree or equivalent practical experience in a technical field Current or pursuing PhD in artificial intelligence or related fields First-authored peer-reviewed conference publications Research background in machine learning or robotics Experience with Python, PyTorch, software, and complex experiments Apply world modelling, predictive learning, Python, and PyTorch research to embodied-agent planning and control at Meta.

📌 Research Scientist World Models (Montreal)
🏢 Meta
📍 Montreal