阿斯塔纳时间
AI QAZAQSTANKZ · AI测试版

哈萨克斯坦人工智能与高科技领域的智能门户

Research1 分钟阅读

内容来自原始来源：

arXiv — cs.AI

在 arXiv — cs.AI 阅读全文

本内容来自公开来源。若发布内容影响您的权利 → 申请删除信息

标题为英语Keep It CALM: Analyzing the Limits of Global Unsafety in Text-to-Image Generation

Training-free safeguards for text-to-image generation often rely on a reusable safety signal, such as an unsafe direction or global toxic subspace, applied broadly across prompts. We provide a controlled geometric analysis of this global-unsafety assumption and reveal a…

arXiv — cs.AI作者: NaHyeon Park, Minhyun Lee, Hyunjung Shim2026年10月5日 (10月5日)
提供 1 种语言版本

AI 摘要：要点

  • Training-free safeguards for text-to-image generation often rely on a reusable safety signal, such as an unsafe direction or global toxic subspace, applied broadly across prompts
  • We provide a controlled geometric analysis of this global-unsafety assumption and reveal a…
Research
KZ·AI 20A7
Инсайт дня

Keep It CALM: Analyzing the Limits of Global Unsafety in Text-to-Image Generation

标签：#arXiv
Поделиться:

相关文章

返回列表