Publications

Beyond Single-Step Likelihood: Gibbs Variational Last Layers for Long-Horizon Dynamics Learning

Paul Brunzema, Henrik Hose, Sebastian Trimpe

Advances in Neural Information Processing Systems (NeurIPS)

Efficient Knowledge Transfer in Federated Bayesian Optimization through Neural Network Surrogates

Alexander Gräfe, Max van Gemmeren, Paul Brunzema, Sebastian Trimpe

Advances in Neural Information Processing Systems (NeurIPS)

Heteroscedastic Variational Last Layers

James Harrison, John Willes, Mikkel Jordahn, Paul Brunzema, Jasper Snoek

Advances in Neural Information Processing Systems (NeurIPS)

Extended version of the AABI 2025 workshop paper. · Collaboration with Google DeepMind, Vector Institute, and DTU

RLLBC-Lib: An Educational Code Library for Reinforcement Learning and Learning-Based Control

Bernd Frauenknecht, Emma Cramer, Artur Eisele, Paul Kruse, Lukas Kesper, Ramil Sabirov, Jyotirmaya Patra, Jonas Hertrampf, Johannes Berger, Paul Brunzema, Friedrich Solowjow, Sebastian Trimpe

European Workshop on Reinforcement Learning (EWRL)

A Decade of Bayesian Optimization for Controller Tuning and Robot Learning: Tutorial, Review, and Future Prospects

David Stenger, Paul Brunzema, Johanna Menn, Alexander von Rohr, Angela P. Schoellig, Sebastian Trimpe

arXiv preprint arXiv:2609.09403

Collaboration with TUM and UTN

Agentic Bayesian Optimization through Surrogate-Augmented Autoresearch

Paul Brunzema, Louis C. Tiao, Nhat Le, Kevin De Angeli, Yao Xuan, Djordje Gligorijevic

arXiv preprint arXiv:2608.00316

Currently under review · Research internship at Meta

Local Preferential Bayesian Optimization

Johanna Menn, Miriam Kober, Paul Brunzema, David Stenger, Sebastian Trimpe

arXiv preprint arXiv:2606.02351

Currently under review

Vision-Conditioned Variational Bayesian Last Layer Dynamics Models

Paul Brunzema, Thomas Lew, Ray Zhang, Takeru Shirasawa, John Subosits, Marcus Greiff

IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS)

Research internship at Toyota Research Institute (TRI)

The Mini Wheelbot Dataset: High-Fidelity Data for Robot Learning

Henrik Hose, Paul Brunzema, Devdutt Subhasish, Sebastian Trimpe

arXiv preprint arXiv:2601.11394

Fine-Tuning of Neural Network Approximate MPC without Retraining via Bayesian Optimization

Henrik Hose, Paul Brunzema, Alexander von Rohr, Alexander Gräfe, Angela P. Schoellig, Sebastian Trimpe

International Conference on Robot Intelligence Technologies and Applications (RiTA)

BayeSQP: Bayesian Optimization through Sequential Quadratic Programming

Paul Brunzema, Sebastian Trimpe

Advances in Neural Information Processing Systems (NeurIPS)

Bayesian Optimization via Continual Variational Last Layer Training

Paul Brunzema, Mikkel Jordahn, John Willes, Sebastian Trimpe, Jasper Snoek, James Harrison

International Conference on Learning Representations (ICLR)

Collaboration with Google DeepMind, Vector Institute, and DTU

Heteroscedastic Variation Bayesian Last Layers

James Harrison, John Willes, Paul Brunzema, Jasper Snoek

Advances on Approximate Bayesian Inference (AABI) Workshop

Collaboration with Google DeepMind, Vector Institute, and DTU

Event-Triggered Time-Varying Bayesian Optimization

Paul Brunzema, Alexander von Rohr, Friedrich Solowjow, Sebastian Trimpe

Transactions on Machine Learning Research (TMLR)

Variational Last Layers for Bayesian Optimization

Paul Brunzema, Mikkel Jordahn, John Willes, Sebastian Trimpe, Jasper Snoek, James Harrison

NeurIPS Workshop on Bayesian Decision-making and Uncertainty

Event-Triggered Safe Bayesian Optimization on Quadcopters

Antonia Holzapfel*, Paul Brunzema*, Sebastian Trimpe

Learning for Dynamics and Control Conference (L4DC)

Neural Processes with Event Triggers for Fast Adaptation

Paul Brunzema*, Paul Kruse*, Sebastian Trimpe

Learning for Dynamics and Control Conference (L4DC)

On Controller Tuning with Time-Varying Bayesian Optimization

Paul Brunzema^*, Alexander Von Rohr*, Sebastian Trimpe

Conference on Decision and Control (CDC)