引导式测验 · 硬 等级

Asking LLMs How Confident They Are Quiz

Tests confidence calibration, ECE, study differences, token likelihoods, task-specific validation, and repeated-answer limits for consequential workflows.

相关引导路径Verbalized Confidence From LLMs
问题 1 的 8

What does a model-generated “80% confident” statement establish by itself?