PromptAudit: Auditing Prompt Sensitivity in LLM-Based Vulnerability Detection 事件

PromptAudit: Auditing Prompt Sensitivity in LLM-Based Vulnerability Detection arXiv:2605.24171v1 Announce Type: cross Abstract: Large language models are increasingly used for vulnerability detection, yet their reliability under different prompt formulations remains uncharacterized. We present PromptAudit, a controlled evaluation framework that isolates prompt effects by fixing the dataset, decoding, and parsing while varying only the prompting strategy. Using five prompting strategies across f