牛哥精选 · 所有

📋 全部 🤖 AI·大模型 ⚡ 效率工具 📝 深度技术 🚀 产品观察 💰 商业科技 🔓 开源项目 🎨 设计创意 📖 阅读推荐 🏷 资源合集 🌱 成长效率

🤖 AI·大模型 arXiv NLP 2026-07-28

Tailored untruths: How personalisation challenges LLM safeguards

个性化如何绕过LLM安全防线？这篇论文揭示了危险漏洞。

arXiv:2510.12993v3 Announce Type: replace Abstract: Large Language Models (LLMs) can generate highly persuasive disinformation, yet little is known ab…

llm安全个性化对抗安全 arxiv论文

🤖 AI·大模型 Dev.to 2026-06-09

Claude Opus 4.8, GPT-5.5, and Gemini 2.5 Pro All Score Exactly 50% on Adversarial Safety — The Frontier Ceiling Is Real

五大前沿AI模型在对抗安全测试中齐刷刷得50%，揭示AI安全天花板真实存在。

Every frontier model provider tells you their latest model is "the safest yet." So I tested that claim — on the newest models money can buy. I ran Cla…

claude opu gpt-5.5 gemini 2.5 对抗安全模型基准测试

📅 日期

2026-05-20 2026-05-19

🐂 牛哥精选

Tailored untruths: How personalisation challenges LLM safeguards

Claude Opus 4.8, GPT-5.5, and Gemini 2.5 Pro All Score Exactly 50% on Adversarial Safety — The Frontier Ceiling Is Real

📅 日期