A Tale of Two Error Categories: Exploring Concealed Trade-Offs in the Errors of Automated Judges in Evaluation of Uncertainty Quantifiers

Key points

• The wide adoption of LLMs across broad NLG applications heightens the importance of providing users with the means to avert errors and hallucinations.1
• Uncertainty quantification is poised to fill that gap; with low uncertainty (high confidence), as a proxy for correctness, allowing users to be selective (e.g., reject low-confidence, likely incorrect responses).1
• Correlation between confidence and correctness then serves as a useful criterion for evaluation of uncertainty quantifiers (UQs).1
• With experiments in question answering, using 4 LLMs, human and automated judgements and 7 popular UQs, we find that i) a judge's performance can only coarsely predict the observed impact of its errors on the reliability of UQ evaluation, and that ii) judgement errors tend to misrepresent informative UQs most.1

Sources (1)

• [1]A Tale of Two Error Categories: Exploring Concealed Trade-Offs in the Errors of Automated Judges in Evaluation of Uncertainty Quantifiers
  arXiv (AI, ML, NLP, CV, robotics, multi-agent) · Oct 7, 08:52 AM

  In a meta-analysis of published work, we show that automated judgement is the present norm.

  The wide adoption of LLMs across broad NLG applications heightens the importance of providing users with the means to avert errors and hallucinations.

Extractive summary: sentences quoted from the sources.