This research explores how Large Language Models can automatically extract failure-inducing inputs from bug reports, streamlining the debugging process for developers.

• LLMs show promising capability in parsing natural language bug reports to identify critical inputs
• The approach automates a traditionally manual and time-consuming security analysis process
• Results demonstrate how AI can bridge the gap between natural language descriptions and technical debugging
• Offers practical applications for improving software security workflows

By automating input extraction, security teams can accelerate vulnerability remediation and reduce the manual effort required for bug triage and analysis.

LLPut: Investigating Large Language Models for Bug Report-Based Input Generation