Calibrating the Confidence of Large Language Models by Eliciting Fidelity

Open in new window