Ars Technica:AI·· 2026-05-29AI 评分70
LLMs相信虚假陈述,即使明确警告也无效
LLMs believe false statements even after explicit warnings that they're false
AI 导读
微调测试显示,大语言模型存在一种偏见,即倾向于自信地将虚假陈述表述为真实,即使明确告知其陈述为假。
来源:Ars Technica:AI · arstechnica.com
LLMs believe false statements even after explicit warnings that they're false
微调测试显示,大语言模型存在一种偏见,即倾向于自信地将虚假陈述表述为真实,即使明确告知其陈述为假。
来源:Ars Technica:AI · arstechnica.com