Federal Grant Award Summary

Yale University received a $596,523 Project Grant from the National Science Foundation's Division of Computing and Communication Foundations (CFDA 47.070 - Computer and Information Science and Engineering) beginning June 1, 2026 and concluding May 31, 2031. This CAREER award supports research and development of next-generation artificial intelligence (AI) foundation models that utilize non-Euclidean representation learning and adaptive curvature in embedding spaces. The primary deliverable is a novel geometric approach incorporating a non-Euclidean graph transformer architecture and learnable curvature estimation methodology designed to more accurately capture complex topological relationships inherent in real-world network data, such as biological networks and molecular interactions.

The research initiative addresses critical limitations in current AI systems' ability to represent complex network-like relationships across domains including drug discovery, biological network analysis, and scientific literature retrieval. Beyond core research outputs, the project includes development of new university curriculum in modern multimodal foundation models and trustworthy AI, outreach programs engaging high school through graduate students, and the creation of more explainable and trustworthy AI systems suitable for high-stakes applications such as healthcare. The enhanced interpretability and reliability of these models is expected to accelerate scientific discovery, improve drug development processes, and advance understanding of age-related diseases while strengthening AI safety through increased system transparency.