Jason Li

Jason Li

Publications

Conference Papers

Explainable Reaction Atlas

Published in IEEE Workshop on Topology meets Artificial Intelligence (TopoInVis Connect), 2026

This paper presents a framework leveraging human-AI collaboration to help users analyze and interpret chemical reaction latent spaces

Recommended citation: Jason Li, Rodrigo de la Nuez Moraleda, Dhruv Meduri, Kelin Zia, and Bei Wang. (2026). "Explainable Reaction Atlas." IEEE Workshop on Topology Meets Artificial Intelligence (TopoInVis Connect) at IEEE VIS. Accepted.
Download Paper

Multi-Agent Board Game Strategy Through Simulation

Published in Interservice/Industry Training, Simulation and Education Conference (I/ITSEC), 2025

This paper investigates multi-agent reinforcement learning for Settlers of Catan, combining graph neural networks with exploration, replay-sampling, and ensemble strategies to address stochastic gameplay, long decision sequences, and negotiation among competing agents. The models are evaluated against random, value-function, game-theoretic, and human baselines, offering broader insights for complex decision-support simulations.

Recommended citation: Andres Espinosa, Cody Flynn, Jason Li, Han Mach, Maximillian Banach, Cathy Quan, Jorg Peters, and Brian Stensrud. (2025). "Multi-Agent Board Game Strategy Through Simulation." Interservice/Industry Training, Simulation and Education Conference (I/ITSEC).
Download Paper