Study shows large language models misuse confidence markers

Original: Can LLMs Use Linguistic Uncertainty Markers to Reliably Reflect Intrinsic Confidence?

Source: arxiv.org ↗