The latter formed the basis of a paper in which Philipp and his colleagues investigated the extent to which reinforcement learning, a specific subtype of machine learning, can be used to design and optimize quantum circuits. In reinforcement learning, a so-called agent – an independently acting computer program – learns to solve a problem by interacting with its environment. In Philipp's case, the problem is to find a quantum circuit that transforms an initial quantum state, in which a computing register of qubits is present, into a specified final state. “The agent can perform actions. In our case, it applies and combines certain quantum gates. By perceiving the change caused by its actions, experiences are generated,” explains the computer scientist. Ultimately, the agent can learn through trial and error which combination of gates is best suited to achieve the desired final state.

Philipp did not plan far in advance to conduct research in the field of quantum computing as a computer scientist. “I think the fundamental way computer science works, the analytical thinking, the problem solving, has always suited me.” However, what particularly appealed to him about the bachelor's degree program in media informatics, which he ultimately chose, was its versatility. In the course on human–computer interaction, for example, students also learned the basics of psychology. “I've always found that an exciting topic.”

“I believe that the exciting topics always lie precisely where you go beyond what is given to you.”

Although Philipp switched to pure computer science for his master's degree, it was still important to him to be able to try out different things. “I believe that the exciting topics always lie precisely where you go beyond what is given to you.” Not only the extensive range of courses offered at LMU, but also various working student positions and his volunteer work at the Kulturspektakel Gauting were all opportunities for Philipp to try new things and learn new skills. During a lecture, the young computer scientist got his first taste of quantum computing. At that point, however, he had not yet considered pursuing this topic in his doctoral studies. He was particularly interested in the machine learning courses, especially reinforcement learning, at his current department. Once it was clear that Philipp would remain in the department as a doctoral student, the only thing he knew for sure was that he wanted to continue researching in this field. Since the department had successfully applied for several quantum computing projects at that time, this topic was eventually brought to his attention. Looking back, he is very happy with this decision. He likes that his doctoral thesis topic is so versatile. He finds interdisciplinary contexts, in which new aspects constantly emerge, particularly appealing, says Philipp. “I am especially motivated by topics that offer room for new perspectives.”

The first Annual Meeting of Munich Quantum Valley was also a new experience for the young computer scientist, who has attended every year since. It is always exciting because so many scientists from different disciplines come together, says Philipp. “I always find it a great source of inspiration.” However, he admits that being confronted with so many presentations outside his field of expertise takes some getting used to. “I'm completely out of my depth when it comes to what exactly happens physically at the hardware level. I never learned or studied that,” Philipp emphasizes with a laugh. Over the years, though, you get used to the way scientists from other disciplines talk about things, he continues, you understand the implications of each other's research results. “You see the consequences for your own work and where the disciplines intersect.” To see the contribution one is making to such a large project, was also nice. “My background has always been in machine learning. I've always had this perspective: ‘Hey, you can make a difference here with machine learning.’” Reinforcement learning in particular has become an established approach in the field of quantum circuit design, but also for the construction of quantum error correction codes, the scientist adds.

Philipp has already submitted his doctoral thesis; all that remains is the defense. Seeing what he has accumulated over the years in terms of publications and experience makes up for the inevitable moments of frustration that are part of a doctorate, says the computer scientist. He would like to continue working in research and is already looking for a new position. Although a period of transition always involves uncertainty, Philipp views it positively – after all, it allows for new opportunities.
 

Published 27 February 2026; Interview 10 December 2025