Project Grant 2617281

The National Science Foundation Division of Mathematical Sciences awarded Temple University $149,980 on September 1, 2026, for a collaborative research project developing a rigorous framework for collaboration between mathematicians and artificial intelligence systems to accelerate mathematical discovery in number theory and combinatorics.

The project establishes and tests a four-phase iterative workflow for mathematical discovery. It generates and curates large datasets of complex number-theoretic and combinatorial objects, applies machine-learning methods to detect non-obvious patterns within those datasets, employs specialized tools to interpret how trained AI models make decisions and translate their internal logic into precise mathematical conjectures and algorithms, and validates resulting claims through rigorous mathematical proof and automated verification. The project will make datasets, software, and educational resources publicly available and will strengthen the mathematics and AI communities through a biweekly virtual seminar, summer schools, conferences, and research opportunities for students and early-career researchers.

Performance takes place in Philadelphia, Pennsylvania, with a period from September 1, 2026, through August 31, 2029. The funding supports the Mathematical and Physical Sciences program (CFDA 47.049), which promotes progress in the mathematical and physical sciences to strengthen the nation's scientific enterprise and enhance understanding of major problems confronting the nation. The project supports national priorities in AI by emphasizing interpretable, trustworthy, and verifiable methods and demonstrates how reproducible methodologies can provide data-intensive scientific fields with reliable tools for pattern discovery.