Nathaniel Keplinger
I’m a Ph.D. student in Computer Science at Vanderbilt University, conducting research at ScopeLab. My core interests are in decision-making under uncertainty, with a focus on reinforcement learning and non-stationary Markov decision processes. I also investigate applying decision theory to cyber-physical systems challenges.
Find me on . Email me at: first name dot s dot last name at vanderbilt dot edu
-
AAMAS 2026 Competition on Evaluating Adaptive Decision Agents under Non-Stationarity:
Concluded the first competition on non-stationarity at the AAMAS Conference 2026. Thank you to all competitors for making the competiton possible! We plan to keep the competition leaderboard open for the rest of the year so if you want to test how your AI algorithms perform under conditions different from those during training, check out https://nsgym.io/.
-
CPS-IoT-Week Tutorial on Decision-Making Under Non-Stationarity:
Co-presented a tutorial on the fundamental concepts, mathematical models, algorithms, and open challenges in decision-making under non-stationarity at IEEE/ACM CPS-IoT week 2026 in Saint-Malo, France.
-
NS-Gym:
One of the developers and maintainers of NS-Gym, a comprehensive, open-source simulation framework for non-stationary Markov decision processes. We got our paper on NS-Gym accepted at NeurIPS 2025! Check out nsgym.io.
-
CPS PI Meeting 2025: NS-Gym Tutorial:
Co-presented a tutorial at the 2025 NSF Cyber-Physical Systems Principal Investigators’ Meeting titled ‘Decision-Making under Non-Stationarity: Concepts, Trends, and Challenges.’
-
Workshop Tutorial @ ICAA 2024:
Gave my first tutorial on NS-Gym at the workshop ‘Recent Trends in Sequential Decision-Making’ during the 2024 International Conference on Assured Autonomy (ICAA).