English(EN) Researchers can now reverse-engineer LLM prompts from output text with near-perfect accuracy

新方法以高精度从输出文本中重建大型语言模型提示

作者 PulseAugur 编辑部 · [1 个来源] · 2026-08-12 17:32

研究人员开发了一种新方法，能够以高精度重建用于从大型语言模型（LLM）生成文本的原始提示。该技术被称为“先前标记预测”，无需访问 LLM 的内部权重即可运行，并且已在各种模型上证明了其有效性。这一发展对使用专有系统提示的组织构成了潜在的安全隐患，因为它可能暴露敏感指令。 AI

影响该技术可能会暴露专有系统提示，对使用大型语言模型的组织构成安全风险。

排序理由详细介绍 LLM 提示重建新方法的学术研究论文。[lever_c_demoted from research: ic=1 ai=1.0]

AI 生成摘要 · Google Gemini · 来自 1 个来源。我们如何撰写摘要 →

本文如何被排名

Signal score: 0 / 100; Composite score across the factors below. Higher = stronger signal that this story matters right now.
Newsworthiness bucket: Tool; 详细介绍 LLM 提示重建新方法的学术研究论文。[lever_c_demoted from research: ic=1 ai=1.0]
Source corroboration: Single-source cluster; Only one publisher covered this so far. Single-source stories can still rank when the publisher is high-authority, but they lack cross-source corroboration.
Topics: safety, model release; Editorial topic classification. Feeds into how the story surfaces on /topic/<slug> hub pages and into the per-entity coverage mix.
AI-industry relevance: High; Clearly on-topic for AI-industry coverage.
Story freshness: 56 days old; Aged out of breaking-news scoring windows; ranking reflects the durable signal from the full source set.

完整方法见我们的编辑标准。

The Decoder TIER_1 English(EN) · Matthias Bastian · 2026-08-12 17:32

研究人员现在能以近乎完美的准确度从输出文本中逆向工程 LLM 提示

<p><img alt="" class="attachment-full size-full wp-post-image" height="768" src="https://the-decoder.com/wp-content/uploads/2026/08/prompt_extracting.png" style="height: auto; margin-bottom: 10px;" width="1376" /></p> <p> Researchers at IIT Bombay and Adobe Research have built an…