Every word counts: A multilingual analysis of individual human alignment with model attention

We carry out this correlation reading (Morger et al., 2022; Eberle et al., 2022; analysis on the participants' respective native Bensemann et al., 2022; Hollenstein and Beinborn, languages (L1) and data from an English experiment 2021; Sood et al., 2020). This approach serves as (L2) of the same participants. We analyse an interpretability tool and helps to quantify the the influence of processing depth, i.e., quantifying cognitive plausibility of language models. However, the thoroughness of reading through the readers' what drives these correlations in terms of differences skipping behaviour, part-of-speech (POS) tags, and between individual readers has not been vocabulary knowledge in the form of LexTALE investigated.