Federal Project Grant Award Summary

New York University received a $209,649 CAREER (Faculty Early-Career Development Program) award from the National Science Foundation's Division of Information and Intelligent Systems under the Computer and Information Science and Engineering program (CFDA 47.070), effective October 1, 2025, with completion targeted for July 31, 2027. This project develops AI systems capable of flexible and robust reasoning in natural language to advance question-answering and fact-checking applications. The core deliverable is a learning-based system that processes textual input through pre-trained neural network models (such as T5) to reformulate text, derive logical conclusions, and verify claims or answers. The system produces human-understandable chains of reasoning through two primary modules: a deduction module that combines statements to generate valid logical inferences, and a verifier module that validates whether derived evidence supports the original claim.

The award addresses a critical limitation in modern AI-powered question-answering systems embedded in search engines and digital assistants—their inability to reliably handle complex queries requiring synthesis and integration of multiple information sources. By developing systems that explicitly demonstrate their reasoning process through natural language output, the project contributes to building trustworthy and auditable artificial intelligence systems. The research is funded in part under the American Rescue Plan Act of 2021 and aligns with the NSF's broader portfolio emphasis on AI reliability, safety, and interpretability across computing applications.