Pushing the Boundaries

Advancing AI for a Peaceful and Humane Future

We strive to create AI that not only grasps human intentions and the physical world but does so in alignment with human values. Our work prioritizes robustness, trustworthiness, and ethical AI development, ensuring these technologies foster a more peaceful and humane future.

Our Team

We are the TRESP Lab at the Institute of Informatics at LMU Munich directed by Prof. Dr. Volker Tresp. TRESP stands for Trustworthy, Responsible, Ethical, Substainable, Peaceful.

We push the limits of AI by developing advanced representation learning techniques that enable machines to understand and reason about the world in a structured and interpretable way.

SEE MORE

Research Topics

Our research spans knowledge graphs, multimodal learning, and AI-driven understanding, allowing systems to seamlessly integrate diverse information sources - text, images, and video - into a comprehensive and actionable model of complex environments.

We are always looking for motivated students to conduct guided research and thesis.

JOIN US

News

Sep 24, 2026 5 Papers Accepted by NeurIPS 2026
Feb 23, 2026 🎤 Prof. Volker Tresp co-organized the Dagstuhl Seminar on Tensor Factorizations Meet Probabilistic Circuits
Jan 30, 2026 We are featured in MCML 2025 Year Report!
Jan 30, 2026 We are co-organizing the ICLR Workshop on Memory for LLM-Based Agentic Systems
Jan 29, 2026 🎤 Prof. Volker Tresp Featured in 2025 AI Progress Report
Jan 28, 2026 In 2025, we have papers accepted at NeurIPS (including one Spotlight), EMNLP, COLM, ACL, CVPR, AAAI (including one oral), ICLR.
Dec 1, 2025 We are looking for HiWis on the research project on Surgical Video Understanding.
Nov 30, 2025 🔥 We have two PhD openings via METEOR and MLCARE projects.
Nov 24, 2025 🏆 Prof. Tresp was selected among the Leading 30 German AI Experts & Academics 2025.
Oct 30, 2025 🏆 TRESP Lab, in collaboration with Torr Vision Group, has ranked among the Top 20 Worldwide in OpenAI's Red-Teaming Challenge!