David Goll

Contribute to the documentation of the institute's high-performance computer cluster, among other tasks.

"Analysis of Multi-Agent Reinforcement Learning from a Statistical Physics Perspective". Developed a novel deterministic approximation of multi-agent independent Q-learning, identifying critical limitations in existing methodologies and demonstrating how complex emergent reinforcement learning dynamics manifest even in elementary multi-agent systems.

Specialised on statistical physics, nonlinear dynamics and computational physics.

Thesis: "Nonlinear Optical Response of Organic Thin Molecular Films Studied with Angle-Resolved Second Harmonic Generation" Parts of the results were published in Nefedov et al. 2021, Advanced Materials.

I had the opportunity to attend a three-day workshop at Princeton, organised by the 'Cooperation and Collective Cognition Network' in collaboration with Humboldt University. There, I presented my research and contributed to discussions on modelling approaches for multi-agent systems.