Position Title: Research Intern

Contract Type: Internship

Educational Requirement: Master's degree or equivalent

Role Overview: The selected candidate will engage in research activities under the guidance of their supervisor. The focus will be on exploring the theoretical aspects of two-layer neural networks with ReLU activation, particularly examining how the stochastic nature of Stochastic Gradient Descent (SGD) influences training dynamics.

About the Research Center:

The Inria research center in Saclay, established in 2008, plays a pivotal role in the development of the Saclay plateau. It collaborates closely with Paris-Saclay University and the Polytechnic Institute of Paris, having formed strategic partnerships to enhance its research capabilities.

Key Responsibilities:

• Conduct a comprehensive literature review
• Implement numerical simulations using Python for neural network models
• Investigate theoretical outcomes related to neural networks through the application of mathematical techniques in non-convex optimization and stochastic differential equations

Required Skills:

• Understanding of neural network theory
• Ability to read and comprehend scientific literature in English
• Proficiency in coding with Python
• Knowledge of non-convex optimization
• Familiarity with differential equations