Curriculum Vitae
Patrick Wienhöft completed his bachelor’s degree at TU Dresden before specializing into temporal logic and verification in the Master’s programme Computational Logic. After the defense of his Master thesis “Human-inspired Reinforcement Learning” he joined CeTI in December 2020 where he continued researching on this topic, combining the mathematical and logical background obtained in his studies and native to his chair with the neuroscientific expertise in CeTI’s other research groups.
Projects/Cooperation within CeTI you are involved in:
My main research area of habitual reinforcement learning offers possibilities for cooperations with psychological and neuroscientific research to develop models that formalize habitual behaviour in humans and transfer the insights to machine learning. Thus the development was influenced by concepts introduced to me by the chair of neuroimaging at TUD.
For the future further projects are planned to solidify the psychological theory behind the approach, as well as cooperations with reasearchers working on explainability in AI. Once I am confident in the learning model I will reach out to the U-rooms for potential applications.
What do you value most about your work at CeTI?
Clearly the interdisciplinarity. I have always had interest for all kinds of research areas and CeTI allows me to incorporate some of them into my own research.
What was your best moment at CeTI so far?
Since I have joined CeTI in this… special time my contacts were mainly limited to online calls. I think finally getting to see many of the faces in real life soon at the summer school and/or GA will be my best moment by far soon.
What else would you like to research?
Linguistics. This is not too far from CeTI’s scope so I can find a possibility to work on this in the future.
Other than that: Speleology. But unfortunately I only discovered my interest in caves only when I already had my degree in computer science so I doubt I will research in that area any time soon.
How do you spend your spare time?
I like to be active and in motion when I’m outside, e.g. by climbing, geocaching, caving, scuba diving, or snowboarding. When at home I spend most of my time gaming.
Publications
| 1. | Lazy model checking for recursive state machines (Journal Article) In: Software and System Modeling, vol. 23, no. 2, pp. 369–401, 2024. |
| 2. | Towards a formal account on negative latency (Book Section) In: Steffen, Bernhard (Ed.): Bridging the Gap Between AI and Reality, vol. 14380, pp. 188–214, Springer, 2023. |
| 3. | More for less: Safe policy improvement with stronger performance guarantees (Proceedings Article) In: Proceedings of the International Joint Conference on Artificial Intelligence (IJCAI), 2023. |
| 4. | Strategy synthesis in Markov decision processes under limited sampling access (Book Section) In: Rozier, Kristin Y.; Chaudhuri, Swarat (Ed.): NASA Formal Methods, vol. 13903, pp. 86–103, Springer, 2023. |
| 5. | Be lazy and don’t care: Faster CTL model checking for recursive state machines (Book Section) In: Calinescu, Radu; Păsăreanu, Corina S. (Ed.): Software Engineering and Formal Methods, vol. 13085, pp. 332–350, Springer, 2021. |
