Paul Krogmeier

Paul Krogmeier is a postdoctoral fellow at Harvard University's School of Engineering and Applied Sciences (SEAS), advised by Nada Amin and Walter Fontana. His research bridges programming languages, logic, and program synthesis with applications in interpretable machine learning and computational universality. Starting Fall 2026, he will join the University of Colorado Boulder as an assistant professor in computer science.

His educational background includes a PhD from the University of Illinois under Madhusudan Parthasarathy.

Dr. Krogmeier investigates computational universality under errors, formal expressivity of neural networks, and symbolic language emergence. His work connects evolutionary processes with learning mechanisms and rigorously characterizes creativity through computational models. Ongoing projects focus on DSL synthesis for few-shot learning and developing frameworks where symbolic language emerges from computational processes.

His publication record shows consistent contributions to programming languages theory (OOPSLA, POPL) with increasing integration of machine learning techniques. Key themes include program synthesis, logic-based learning, and formal analysis of neural architectures, demonstrating interdisciplinary rigor across theoretical computer science and AI.

His scientific achievements include:

• Distinguished Paper Award at OOPSLA 2023
• Distinguished Paper Award at POPL 2022

While currently not advising students as a postdoc, he will establish an advising program at CU Boulder. His research is supported by institutional resources at Harvard SEAS with potential future funding from NSF and computing industry partnerships.

At Harvard, he collaborates within Nada Amin and Walter Fontana's research group, which explores foundational programming language concepts applied to machine learning systems and biological computation models.