Project Grant 2403318

This National Science Foundation (NSF) Project Grant award under the Computer and Information Science and Engineering (CFDA 47.070) program provides $316,000.00 to The Leland Stanford Junior University (Stanford University) to develop a general foundation model framework for handling graph-structured data in scientific discovery. The key products and services to be delivered include:

Addressing limitations in existing graph foundation models by incorporating novel approaches such as multi-level graph neural networks, graph signal processing, multimodal graph learning, graph-specific fine-tuning, and in-context learning. This framework aims to capture human scientific knowledge and express the complexity of the natural world to dramatically transform machine learning models in scientific discovery. The framework will enable tackling a wide range of complex scientific tasks, even with scarce supervision labels. The project has an ultimate completion date of July 31, 2028.