Skip to content
Hot eventLive

FDPR:面向LLM漏洞分析的失败驱动提示词优化

1 reports1 sources3 hr ago updated

Get the story

AI overview

2026年10月7日,arXiv Software Engineering 发表研究,提出 Failure-Driven Prompt Refinement(FDPR),通过分析大语言模型在漏洞分析中的反复失败来指导提示词优化。研究基于 Damn Vulnerable Java Application(DVJA)识别出误报、漏报、推理无依据和 CWE 误分类等失败模式,并在 Juliet Test Suite 上进行评估与跨模型验证。该方法旨在从失败案例中提炼改进方向,提升 LLM 漏洞分析的准确性。

Generated from reports · updated 3 hr ago

Timeline

Follow the coverage from different angles.

Oct 7, 2026
  1. arXiv · Software Engineering
    Learning from Failures:面向 LLM 漏洞分析的失败驱动提示词优化

    该研究提出 Failure-Driven Prompt Refinement(FDPR),通过分析大语言模型在漏洞分析中的反复失败来指导提示词优化。研究基于 Damn Vulnerable Java Application(DVJA)识别出误报、漏报、推理无依据和 CWE 误分类等失败模式,并在 Juliet Test Suite 上评估与跨模型验证。

Heat trend

Current heat 9·Comparable peak 10(Oct 7)·Comparable change over 24 hours –

02.557.510Oct7Oct7Oct7Oct7

The trend compares only the same participants observed continuously; its range may be smaller than the current heat count. Move or click on the chart to inspect hourly heat; use the left and right arrow keys to switch.