R&D Intern - Fairness in Deep Learning - Paris
Core
Research internship on ensuring prediction independence from continuous sensitive variables (e.g., age, income, skin color) in neural networks.
Role type
Research Intern (Master's level)
Builds
Robust protocols for comparing bias reduction methods and new differentiable debiasing techniques for continuous variables.
Domain
Artificial Intelligence / Machine Learning / Algorithmic Fairness
Deliverable
research
Required skills
Neural networks, continuous variable analysis, dependency measurement, PyTorch, scikit-learn
Preferred skills
Mathematical modeling, statistical learning, experimental design
Technologies
PyTorch, scikit-learn
Responsibilities
Implement existing bias reduction methods (e.g., Hirschfeld-Gebelein-Rényi correlation, Generalized Disparate Impact), design a robust comparison protocol, develop new differentiable debiasing methods for continuous sensitive variables.
Seniority
Intern (Master's level)
