Rui Yuan

I am currently an AI research scientist at Stellantis. I received my PhD in Applied Mathematics from a joint program between Meta AI (prev. Facebook AI Research) and Télécom Paris, advised by Alessandro Lazaric (Meta AI) and Robert M. Gower (Flatiron Institute).

Research

My research interests include reinforcement learning (RL), optimization, deep learning, and more recently, reinforcement learning from human feedback and in-context learning.

During my PhD, I worked on RL and large-scale stochastic optimization for machine learning. My focus centered on gaining a fundamental understanding of the optimization methods employed in RL and designing new efficient practical optimization algorithms to solve RL problems.

Publications

Simone Rossi, Rui Yuan, Thomas Hannagan. Understanding In-Context Learning in Transformers, International Conference on Learning Representations (ICLR) Blogposts Track, 2024.
Blog Cite

Carlo Alfano, Rui Yuan, Patrick Rebeschini. A Novel Framework for Policy Mirror Descent with General Parameterization and Linear Convergence, Neural Information Processing Systems (NeurIPS), 2023.
PDF Cite Poster Slides NeurIPS

Rui Yuan. Stochastic Second Order Methods and Finite Time Analysis of Policy Gradient Methods, Thesis, 2023.
PDF Cite Slides Thesis

Rui Yuan, Simon S. Du, Robert M. Gower, Alessandro Lazaric, Lin Xiao. Linear Convergence of Natural Policy Gradient Methods with Log-Linear Policies, International Conference on Learning Representations (ICLR), 2023.
PDF Cite Poster Slides ICLR

Rui Yuan, Robert M. Gower, Alessandro Lazaric. A General Sample Complexity Analysis of Vanilla Policy Gradient, International Conference on Artificial Intelligence and Statistics (AISTATS), 2022.
PDF Cite Poster Slides AISTATS

Jiabin Chen*, Rui Yuan*, Guillaume Garrigos, Robert M. Gower. SAN: Stochastic Average Newton Algorithm for Minimizing Finite Sums, International Conference on Artificial Intelligence and Statistics (AISTATS), 2022.
PDF Cite Code Poster Slides Video AISTATS

Rui Yuan, Alessandro Lazaric, Robert M. Gower. Sketched Newton-Raphson, Society for Industrial and Applied Mathematics (SIAM) Journal on Optimization (SIOPT), 2022.
PDF Cite Slides Video SIAM

Misc

I hold an MSc/BSc in Applied Mathematics from École Polytechnique (diplôme d'Ingénieur) and an MSc in Machine Learning from Université Paris-Saclay (Master Data Sciences). During my academic journey, I interned in Machine Learning at Télécom Paris and served as a teaching assistant at the African Institute for Mathematical Sciences (AIMS) in Kigali, Rwanda, with Prof. Robert M. Gower.
Outside of work, I enjoy playing basketball🏀, badminton🏸 and skiing⛷.
As an animal enthusiast, I have a soft spot for dogs and cats. I'm particularly fond of golden retrievers🐕 and orange cats🐈.