LLMs and humans both struggle with machine translation quality annotation
Original titleLarge Language Models for Machine Translation Quality Annotation: Humans and Models Are Both Challenged
AISummary
Apple Machine Learning Research evaluates large language models against human annotators on Multidimensional Quality Metrics (MQM) and Error Span Annotation (ESA), two machine translation quality schemes, across a 70-language-pair test set and WMT23 and WMT25 data.
Source: Apple Machine Learning Research · machinelearning.apple.comPublished · added here