跳转至

GPT-4 Technical Report

约 111 个字 预计阅读时间 1 分钟

由两篇论文组成,介绍了GPT-4相比与之前的模型的强大,以及重点关注GPT-4给出错误/危险/违法/编造回答的可能,给出了各种错误的示例与尝试减少错误的方法及效果。

错误类型有:

  • Hallucinations
  • Harmful content
  • Harms of representation, allocation, and quality of service
  • Disinformation and influence operations
  • Proliferation of conventional and unconventional weapons
  • Privacy
  • Cybersecurity
  • Potential for risky emergent behaviors
  • Interactions with other systems
  • Economic impacts
  • Acceleration
  • Overreliance

评论