新闻
When Should LLMs Abstain? Chain-of-Self-Questioning for Selective Risk Control
arXiv cs.AI · 发布于 · 阅读约3分钟
30秒读懂
- 发生了什么
- Chain-of-Self-Questioning framework lets LLMs decide whether to answer or abstain by assessing information sufficiency before committing to responses.
- 为何重要
- Matters for engineers building systems where wrong answers are costly, like medical advice, legal guidance, or safety-critical applications requiring reliable refusal.
- 注意
- Framework tested only on multiple-choice and short-answer formats; unclear how well self-assessment transfers to open-ended generation or real-world deployment scenarios.
- llm
- language model
- prompt
- eval
- open-weight
这条新闻背后的模式
每个模式都讲清楚技术如何运作、何时值得投入，以及在哪里会失效。
The Agent Architect
每周一个模式、一个权衡、一个生产事故案例。为构建智能体系统的人准备的每周简报。
每周一封邮件,一键退订。您的地址仅用于发送简报。