EUNO.NEWS EUNO.NEWS
  • All (2364) +206
  • AI (546) +17
  • DevOps (142) +2
  • Software (996) +129
  • IT (675) +57
  • Education (5) +1
  • Notice
  • All (2364) +206
    • AI (546) +17
    • DevOps (142) +2
    • Software (996) +129
    • IT (675) +57
    • Education (5) +1
  • Notice
  • All (2364) +206
  • AI (546) +17
  • DevOps (142) +2
  • Software (996) +129
  • IT (675) +57
  • Education (5) +1
  • Notice
Sources Tags Search
한국어 English 中文
  • 3天前 · ai

    AI模型阻止了87%的单次攻击,但在攻击者持续攻击时仅阻止8%

    一个恶意提示被拦截,而十个提示成功通过。这个差距决定了通过基准测试与抵御真实世界攻击之间的区别。

    #adversarial attacks #prompt injection #LLM security #model robustness #enterprise AI
  • 1周前 · ai

    [论文] 注意力引导的 Patch-Wise 稀疏对抗攻击在 Vision-Language-Action 模型上

    近年来,具身智能中的 Vision-Language-Action(VLA)模型发展迅速。然而,现有的 adversarial attack 方法需要高成本……

    #adversarial attacks #vision-language models #embodied AI #feature-space perturbation #multimodal robustness
EUNO.NEWS
RSS GitHub © 2025