We extend reading goal decoding to open-ended information seeking, showing that specific search targets can be identified from eye movement patterns during reading.

A comprehensive benchmark for predictive modeling of human reading behavior from eye movements, featuring standardized tasks, datasets, and evaluation protocols.

We show that the goal of reading — ordinary reading vs. information seeking — can be decoded from eye movements, and that reading goal has a substantial impact on eye movement behavior.

This study investigates the challenge of predicting individual reading comprehension on a single passage level from eye movements using large-scale eyetracking data.