Anas Barakat
Research Fellow
Singapore University of Technology and Design
My research develops foundations and algorithms for sequential and strategic decision-making when objectives extend beyond expected cumulative reward, feedback is limited, and multiple adaptive agents interact. My work lies at the intersection of optimization, reinforcement learning, and learning in games, with a particular interest in emerging problems in modern AI, including post-training, AI alignment, and multi-agent learning.
I am currently a Research Fellow at Singapore University of Technology and Design, working with Georgios Piliouras and Antonios Varvitsiotis. Previously, I was a postdoctoral fellow at ETH Zurich working with Niao He. I received my PhD in mathematics and computer science from Institut Polytechnique de Paris (Télécom Paris), advised by Pascal Bianchi and Walid Hachem.
Research
Optimization for Sequential and Strategic Decision-Making
My research develops foundations and algorithms for decision-making beyond classical assumptions on objectives, feedback, and agent interactions, spanning three complementary directions:
-
Policy optimization beyond expected additive rewards: I study reinforcement learning and policy optimization when objectives depend nonlinearly on distributions of states, trajectories, or model responses, including general-utility RL and modern post-training.
-
Learning in strategic and dynamic environments: I study decentralized learning in nonstationary environments where multiple agents interact using game theory, including static, Markov and continuous-state dynamical games, with a focus on learning dynamics, stability, and convergence to equilibria.
-
Online learning with hidden structure and limited feedback: I study when learning remains possible despite apparent nonconvexity or restricted feedback. My work exploits latent structure such as hidden convexity and develops algorithms that learn directly from bandit or pairwise preference feedback.
Selected & Recent Research
For a complete list of my publications, see below or my Google Scholar.
News
Experience and Background
Here is my CV for more information.
Research
Publications
Policy Optimization Beyond Expected Additive Rewards
I study policy optimization when objectives depend nonlinearly on the distribution of trajectories or model responses, rather than on expected cumulative reward alone. My work develops scalable and globally convergent methods for general utility reinforcement learning, risk-sensitive objectives, and modern post-training objectives.
Learning in Strategic and Dynamic Environments
I study decentralized learning in strategic environments, from static games to Markov and continuous-state dynamical games. My work uses geometric and game-theoretic structure to understand when learning dynamics converge, cycle, or become unstable, and to design algorithms with equilibrium and regret guarantees.
Online Learning with Hidden Structure and Limited Feedback
I study online learning when favorable structure may be hidden and feedback is limited or indirect. My work asks when latent geometry can restore tractability, when weak feedback—such as pairwise preferences—fundamentally limits what can be identified or learned, and how to design learning objectives and algorithms that remain meaningful under such informational constraints.
Earlier Work: Stochastic Optimization
My earlier work studied stochastic approximation and non-convex optimization, including adaptive methods and dynamical systems perspectives on stochastic algorithms.
Student Mentoring
-
Olivier Lepel — Master’s thesis, Mar.–Sep. 2024
Policy Gradients for Cumulative Prospect Theory in Reinforcement Learning. TMLR 2026.
Currently: Data Scientist at BCG. -
Kimon Protopapas — Master’s semester project, Sep. 2023–Jan. 2024
Policy Mirror Descent with Lookahead. NeurIPS 2024.
Currently: Research Intern at University of Basel. -
Philip Jordan — Master’s thesis, May–Dec. 2023
Independent Learning in Constrained Markov Potential Games. AISTATS 2024.
Currently: PhD student at EPFL. -
Jiduan Wu — Master’s thesis, Oct. 2022–Mar. 2023
Learning Zero-Sum Linear Quadratic Games with Improved Sample Complexity and Last-Iterate Convergence. IEEE Conference on Decision and Control (CDC) 2023, SIAM Journal on Control and Optimization, 2025.
Currently: PhD student at Max Planck ETH Center for Learning Systems. -
Harish Rajagopal — Master’s thesis, Mar.–Sep. 2022
Multistage Step Size Scheduling for Minimax Problems.
Currently: Quantitative Technologist at Qube Research & Technologies. -
Julien Lehmann — Master’s thesis, May–Oct. 2021
Analysis of a Target-Based Actor-Critic Algorithm with Linear Function Approximation. AISTATS 2022.
Currently: Research Engineer at OCamlPro.
Teaching
SUTD (2024-2025):
- Co-instructor and course designer ( Online Learning and Learning in Games) with John Lazarsfeld and Iosif Sakos.
ETH Zurich (2022-2024):
- Teaching Assistant and instructor for 2 lectures for the course Optimization for Data Science (Prof. Niao He).
- Head Teaching Assistant and instructor for 1 lecture (on value-based methods) for Foundations of Reinforcement Learning (Prof. Niao He).
- Coordinator of the seminar course ‘Advanced Topics in Machine Learning’.
Télécom Paris (2018-2021): Teaching Assistant
Optimization for Machine Learning (graduate level), Statistics (graduate level), Probabilities (undergraduate level).
Academic Service
- Area Chair: ICLR 2027.
- Conference Reviewing: NeurIPS (‘22, ‘23, ‘25, Top Reviewer ‘26), ICML (‘25, Gold Reviewer ‘26), ICLR (‘25, ‘26), AISTATS (Top Reviewer ‘22, ‘24, ‘25, ‘26), EC (‘26).
- Journal Reviewing: Journal of Machine Learning Research (JMLR), Transactions on Machine Learning Research (TMLR), Mathematical Programming, SIAM Journal on Optimization (SIOPT), Journal of Optimization Theory and Applications (JOTA), IEEE Transactions on Automatic Control.