Explore physics-inspired neural network compression at Amii, using statistical physics and machine learning to study effective AI systems. Collaborate across research teams and ESA.

As a Research Associate, you will contribute to SPAARC, a project with the European Space Agency. Your work will connect renormalization-group theory, knowledge distillation, neural architecture optimization, and statistical analysis. You will lead experimental research while collaborating with Amii and ESA scientists on computer vision and space-based Earth observation applications.

Key Responsibilities:

- Develop theoretical foundations for renormalization-group approaches to compression
- Formulate mathematical frameworks and testable research hypotheses
- Develop and evaluate compression and distillation algorithms




- Design numerical experiments on scaling and compression limits
- Analyze uncertainty, robustness, and architectural sensitivity

Requirements:

- PhD in Physics or Mathematics preferred
- Strong statistical physics and renormalization-group foundation
- Extensive deep learning design, training, and evaluation experience
- Strong Python and PyTorch proficiency
- Research experience with rigorous quantitative analysis

Apply physics, machine learning, and quantitative research skills to neural network compression for space-ready AI systems.

📌 Physics-Inspired Machine Learning Research Associate (Edmonton)
🏢 Amii (Alberta Machine Intelligence Institute)
📍 Edmonton