29/05/2026
💬 How often do you hear, 'Because I asked AI'? How many people know that AI can hallucinate? Will this issue be solved anytime soon?
At , it’s certainly a work in progress. Congratulations to Dr. Daniel Höller for receiving a €1.8 Million Emmy Noether funding from the German Research Foundation ( ) for his research project, 'Neuro-Symbolic Methods in Sequential Decision Making'! 👏
His work was among the 15 projects (out of 178) that were selected. In his research, Höller combines two approaches in automated planning: symbolic and neural.
Symbolic approaches operate on logic, are more rigid, and have explainable action processes. While this leads to more reliable results, they fail in larger systems. Meanwhile, neural approaches require less data and recognize patterns in large systems, making them useful for real world applications. However, they’re unexplainable and can hallucinate. 🤖
Höller aims to achieve an equilibrium. His Emmy Noether group, 'Neuro-Symbolic Methods in Sequential Decision Making', will work on systems that can provide guarantees despite integrating machine learning.
In today's world where LLM outputs are taken as fact, continues to prioritize reliable and secure AI with achievements like these.
🔗 Read more: sic.link/neurosymbolic
Universität des Saarlandes - Saarland University The Max Planck Institute for Software Systems Fachschaft Informatikstudiengänge / CS Students' Representative Council UdS Deutsches Forschungszentrum für Künstliche Intelligenz