Artificial Analysis· ArtificialAnlys·· 8 小时前AI 评分38
英美英语致12模型性能显著差异
12 of 49 models show statistically significant differences in performance between US and UK English, with non-overlapping confidence inte...
AI 导读
49个模型中有12个在美式和英式英语间的表现存在统计显著差异,置信区间不重叠。7个模型在美式英语中表现更好,而5个更偏向英式英语。所有模型的平均差异仅为7 Elo点,
来源:Artificial Analysis · x.com