Shifting Attention to Relevance: Towards the Uncertainty Estimation of Large Language Models

Open in new window