Genentech: Pan Kessel | Machine Learning Scientist, Prescient Design, Computational Sciences

Scientists / Our Scientists

I obtained my PhD in String Theory from the Max Planck Institute for Gravitational Physics. I then switched to Machine Learning and became a member of the Machine Learning Group at TU Berlin.

Featured Publication

Asymptotically unbiased estimation of physical observables with neural samplers

Phys. Rev. E 101, 023304

Kim A. Nicoli, Shinichi Nakajima, Nils Strodthoff, Wojciech Samek, Klaus-Robert Müller, and Pan Kessel

View Abstract

Estimation of Thermodynamic Observables in Lattice Field Theories with Deep Generative Models

Phys. Rev. Lett. 126, 032001

Kim A. Nicoli, Christopher J. Anders, Lena Funcke, Tobias Hartung, Karl Jansen, Pan Kessel, Shinichi Nakajima, and Paolo Stornati

View Abstract

Explanations can be manipulated and geometry is to blame

Ann-Kathrin Dombrowski, Maximillian Alber, Christopher Anders, Marcel Ackermann, Klaus-Robert Müller, Pan Kessel

View Abstract

Path-Gradient Estimators for Continuous Normalizing Flows

PMLR 162:21945-21959

Lorenz Vaitl, Kim Andrea Nicoli, Shinichi Nakajima, Pan Kessel

View Abstract

My research focus is mainly on generative models and their applications to physics. In particular, I am interested in using generative models to speed up sampling problems. I have also worked on interpretability and its connection to generative models. Recently, I have been also thinking about NTK theory and Bayesian optimization.

Max Planck Institute, PhD – 2017

Attention: This page may not render properly. The browser you are using, Internet Explorer 10, is no longer supported.

Pan Kessel

Machine Learning Scientist, Prescient Design, Computational Sciences

Featured Publication

Asymptotically unbiased estimation of physical observables with neural samplers

Estimation of Thermodynamic Observables in Lattice Field Theories with Deep Generative Models

Explanations can be manipulated and geometry is to blame

Path-Gradient Estimators for Continuous Normalizing Flows