Larkin Liu - επ£²
π€ Bio
My research program is centered on the mathematical foundations of intelligent systems that learn and adapt in complex, uncertain environmentsβfrom competitive supply chains to swarms of robots.
My methodological approach combines rigorous theoretical analysis with high-performance computing. I develop and analyze algorithms for:
- π€ Learning in the Wild: Online learning and bandit algorithms that enable decision-making under incomplete information.
- π§ Strategic Planning: Monte Carlo tree search and simulation-based methods for long-horizon sequential decisions.
- βοΈ Game-Theoretic Intelligence: Equilibrium computation and mechanism design for multi-agent interactions.
β Current Research Focus:
Developing theoretically-grounded algorithms for stochastic multi-agent environments and bridging online learning with game theory for applications in swarm robotics and competitive supply chains.
My work directly informs the design of robust operations management systems (e.g., dynamic pricing, resilient supply chains) and intelligent control systems (e.g., autonomous robotics).
Prior to my doctoral studies, I designed and deployed production-grade machine learning pipelines at scale for eCommerce, legal tech, and ad-tech organizations.
Additional details regarding my research activities and publications are available on my academic website at TU Munich.
π° Recent Updates
-
2026
- Jul 2026 π€ Invited Talk: Incentivisation of Stochastic Games: Contracts & Reductions β Workshop for Junior Researchers in Economics and Computation (JECCO UK 2025), University of Oxford, UK.
- Jul 2026 π Doctoral Defense: Online Learning and Planning for Stochastic Multi-Agent Games β Technical University of Munich, Germany. β
- Jun 2026 π Research Visit: Shanghai Jiao Tong University (Host: Shuai Li) β Shanghai, China.
- Apr 2026 π Attended Workshop: Workshop on Geometric Analysis β Universitat de ValΓ¨ncia, Spain.
- Feb 2026 π Attended Workshop: Learning from Heterogeneous Sources β Simons Institute, Berkeley, USA.
- Jan 2026 π Accepted Paper: Riemannian Manifold Learning for Stackelberg Games with Neural Flow Representations at AAAI 2026.
-
2025
- Nov 2025 π€ Invited Talk: Optimal Incentivation for Multi-Follower Principal Agent Stackelberg Games β BDDAI Workshop, King's College London.
- Oct 2025 π Attended Workshop: 30 Years of Game Theory β Institut Henri PoincarΓ©, Paris, France.
- Oct 2025 π Accepted Paper: Online Mixture of Experts: No-Regret Learning for Optimal Collective Decision-Making accepted to NeurIPS 2025.
- Oct 2025 π€ Invited Talk: Online Mixture of Experts β PRAGMA Fest!, AGH University of KrakΓ³w, Poland.
- Jun 2025 π Attended Summer School: Cooperative AI Summer School β Cooperative AI Foundation, Marlow, UK.
- Jun 2025 π Attended Summer School: 2nd ACM Games and AI Summer School β UniversitΓ© Toulouse Capitole, France.
- Feb 2025 π€ Invited Talk: Advances in Multi-Agent Stochastic Optimization β Workshop on Uncertainty in ML, Milan, Italy.
-
2024
- Oct 2024 π Accepted Paper: No-Regret Learning for Stackelberg Equilibrium Computation at ADT 2024 (Piscataway, USA).
- Sep 2024 π Accepted Paper: Approximate Nash Equilibrium Learning for n-Player Markov Games at EPIA 2024 β π Best Paper Award.
- Aug 2024 π Attended Workshop: Lecture Series: Forward From the Fields Medal β Fields Institute, Toronto, Canada.
-
2023
- Oct 2023 π€ Invited Talk: Large Scale Optimization via Monte Carlo Tree Search β Julia and Optimization Days, CNAM, Paris, France.
-
2022
- Oct 2022 π Organized Workshop: Multidisciplinary Workshop on Stochastic Modelling and Monte Carlo Tree Search β Munich Data Science Institute, Germany.
- Jul 2022 π Attended PhD School: EURO StochMod PhD School β Castle Gimborn, Marienheide, Germany.
Academic Background
- Technical University of Munich (2026)
- PhD in Computer Science
- Advisor: Prof. Jalal Etesami
- Dissertation: Online Learning and Planning for Stochastic Multi-Agent Games
- University of Toronto (2017)
- MASc in Operations Research
- Advisor: Prof. Viliam Makis
- Thesis: Comparative Study between Statistical Fraud Detection Methods on eCommerce Networks
- University of Toronto (2015)
- BASc in Mechanical Engineering, with Honours
- Minor in Robotics & Mechatronics
Awards
- Ergodic AI Research Fellowship (2025)
- Mitacs Accelerate Industry Government Joint Research Grant (2015)
- Wallace G Chalmers Engineering Design Award (2013)
- Faculty of Applied Science Engineering Research Fellowship (2012)
- Magna Family Scholarship (2010)