Every word counts: A multilingual analysis of individual human alignment with model attention
This work addresses the problem of generalizing human-model alignment findings across languages and individuals for researchers in computational linguistics and cognitive science, though it is incremental as it builds on prior correlation studies.
The study analyzed eye-tracking data from speakers of 13 languages reading in their native language and English, finding that individual reading behaviors and vocabulary knowledge significantly influence the alignment between human fixation patterns and Transformer-based attention, with considerable differences between languages.
Human fixation patterns have been shown to correlate strongly with Transformer-based attention. Those correlation analyses are usually carried out without taking into account individual differences between participants and are mostly done on monolingual datasets making it difficult to generalise findings. In this paper, we analyse eye-tracking data from speakers of 13 different languages reading both in their native language (L1) and in English as language learners (L2). We find considerable differences between languages but also that individual reading behaviour such as skipping rate, total reading time and vocabulary knowledge (LexTALE) influence the alignment between humans and models to an extent that should be considered in future studies.