POSIX: Prompt Sensitivity Index for Large Language Models
Despite their remarkable capabilities, Large Language Models (LLMs) are found to be surprisingly sensitive to minor variations in …
PDF Cite Poster Video DOI