This $402,777 federal Project Grant award from the National Science Foundation's (NSF) Computer and Information Science and Engineering (CISE) program (CFDA 47.070) supports research at Purdue University to investigate and address limitations of large language models (LLMs) for safe and reliable program generation. The project aims to advance the understanding of LLM errors in program generation, develop methods to localize and mitigate these errors, and create new lightweight model adaptation techniques to address non-functional concerns in LLM-generated code. The research findings and educational activities are expected to inform the design of instructional materials and pedagogical approaches to prepare computer science students for careers in the age of large language models. The award has a performance period from July 1, 2024 to June 30, 2029. No subawards are planned under this grant.
Generated 2/25/25, 3:07 AM