BriefGPT.xyz
Nov, 2024
交叉同情偏见和理解的定量评估
Quantitative Assessment of Intersectional Empathetic Bias and Understanding
HTML
PDF
Vojtech Formanek, Ondrej Sotolar
TL;DR
本研究针对当前同情心定义松散的问题,提出了一种基于心理学起源的同情心评估框架,以提高数据集质量和模型评估的可靠性。研究发现,通过控制生成的提示中社会偏见的变化,能够有效测量大型语言模型对同情心的理解差异,为未来评估方法的构建奠定了基础。
Abstract
A growing amount of literature critiques the current operationalizations of
Empathy
based on loose definitions of the construct. Such definitions negatively affect dataset quality, model robustness, and evaluation reliability. We propose an
→