arXiv · Statistics Machine Learning· Chenyu Zhang, Rachel Luo, Boyi Li, Anjali Parashar, Marco Pavone, Apoorva Sharma·· 4 hr agoAI score32
CARE：安全高效的人机协同决策框架
Careful Judge: Safe and Efficient Human-AI Collaborative Decision Making
AI brief
论文提出 CARE（calibrated adaptive rectification and escalation）框架，将 AI 模型与人类审核者结合，在保证安全、对齐人类决策的同时持续从人类反馈中学习以减少人工查询。
Source: arXiv · Statistics Machine Learning · arxiv.org