Mark Nelson

Mark Nelson is the Charles Lee Powell Professor at American University's Department of Computer Science, with joint appointments in the interdisciplinary AU Game Center. He holds a PhD from Georgia Institute of Technology and a BSc from Harvey Mudd College. His research program bridges artificial intelligence and games, focusing on algorithmic decision-making robustness, game-playing bots, design-support tools, and formal models of game mechanics.

Dr. Nelson leads the NSF-funded project 'Characterizing Algorithm-Relative Difficulty of Agent Benchmarks' (2020-2024) which investigates scaling properties of AI decision-making algorithms. His group developed MCTSLib, an open-source library for Monte Carlo Tree Search that identified key limitations in real-world applications. Additional research explores computer music, domain-specific languages, and computing history.

His publications demonstrate consistent focus on AI robustness, generative models, and educational applications. Recent works explore social physics games using LLM-based dialogue systems, prompt engineering for procedural content generation, and evolutionary methods for language model optimization. This reflects a trajectory toward increasingly complex human-AI interaction systems.

Honors include the NSF CRII Award supporting his research program initiation. He has mentored graduate students now positioned at Google and health-tech startups. Research occurs through the Computational Game Lab, focusing on agent-based simulation frameworks.