Theoretical physics has long provided powerful frameworks for understanding living matter, from stochastic models of molecular kinetics to continuum descriptions of tissues and collective dynamics. Today, rapid advances in AI, especially foundation models and multimodal learning, are transforming how biological systems are studied. Trained on vast, heterogeneous datasets, AI models reveal structure and make predictions across scales with unprecedented scope. These developments challenge the traditional division between theory and experiment and call for new research perspectives guiding and interpreting AI-driven discoveries.
This symposium will focus on the emerging interface between theoretical physics and AI. While physics emphasizes mechanistic and predictive models, AI excels at extracting high-dimensional patterns without explicit assumptions. Bridging these complementary approaches is an important and timely conceptual frontier. Recent progress in physics-informed machine learning, hybrid modeling, and AI-driven theory discovery points to new ways of identifying relevant degrees of freedom, uncovering emergent behaviors, and testing theoretical principles in complex biological systems. By bringing together theoretical physicists, computer scientists, and biomedical researchers, this workshop aims to define new roles of theory in the AI era and to promote collaborations that combine theoretical physics approaches with modern AI to reveal the organizing principles of living matter.
Organisers:
Maria Colome-Tatche (LMU)
Carsten Marr (Helmholtz, LMU)
Steffen Rulands (LMU)
Antonio Scialdone (Helmholtz)