Rieken Luo

PhD Student, Institute of Science and Technology Austria (ISTA)

Am Campus 1, 3400 Klosterneuburg, Austria · rluo@ist.ac.at

News

In the summer of 2026, I will be working at Microsoft Research Cambridge as a Research Intern in Machine Learning.

Research

My research focuses on the intersection of optimization, games, and machine learning. Specifically, I have been working on:

1. Complexity and algorithms for sequential decision-making problems
   • Linear Equations with Min and Max Operators: Computational Complexity
     AAAI'25 (Oral) [slides] [paper]
     (with Krishnendu Chatterjee, Raimundo Saona, Jakub Svoboda)
   • Algorithms for Linear Equations with Min and Max Operators under (Absolutely) Halting Condition
     Submitted [slides] [paper]
     (with Krishnendu Chatterjee, Raimundo Saona, Jakub Svoboda)
   • Improved Algorithmic Analysis for Stochastic Games
     Submitted
     (with Krishnendu Chatterjee, Raimundo Saona, Jakub Svoboda)
   • Algorithms with Smoothed Polynomial-Time Complexity for Deterministic Discounted-sum and Mean-payoff Games
     In preparation
     (with Ali Asadi, Krishnendu Chatterjee)
2. Efficient protocols for multiplayer games
   • Monotone Near-Zero-Sum Games
     ICLR'26 [poster] [slides] [paper]
     (with Sebastian U Stich, Krishnendu Chatterjee)
   • Efficient Gradient Methods for Distributed Saddle Problems
     Submitted [paper]
     (with Anton Rodomanov, Sebastian Stich)
3. Communication-efficient algorithms for distributed consensus optimization
   • Revisiting LocalSGD and SCAFFOLD: Improved Rates and Missing Analysis
     AISTATS'25 [poster] [slides] [paper]
     (with Sebastian U Stich, Samuel Horvath, Martin Takac)
   • On the Convergence of Local SGD under Third-Order Smoothness and Hessian Similarity
     OPT'23 [paper]
     (with Ali Zindari, Sebastian U Stich)