1
Consistent but Miscalibrated: Evaluating LLM Limitations for Risk Communication in Natural Language
一致却失准：大模型在自然语言风险沟通中暴露校准短板，值得AI从业者警惕。
arXiv:2607.03882v1 Announce Type: cross Abstract: LLMs are increasingly deployed as post-hoc explainers of AI-generated outputs, yet it remains unclea…