CoLAGaze

Corpus of Eye Movements for Linguistic Acceptability

CoLAGaze dataset [Bondar et al., 2025].

This dataset includes eye-tracking data from native speakers of English reading sentences from the CoLA dataset. Eye movements are recorded at a sampling frequency of 2,000 Hz using an EyeLink 1000 eye tracker and are provided as pixel coordinates.

Check the respective paper for details [Bondar et al., 2025].

Warning

This dataset currently cannot be fully processed by pymovements due to an error during parsing of individual files.

See issue #1401 for reference.

How to Download

import pymovements as pm

# Initialize the dataset object with its name
# Specify your local directory for saving and loading data
dataset = pm.Dataset(name='CoLAGaze', path='path/to/your/data/directory')

# Download the dataset and extract all archives.
dataset.download()

# Load the dataset into memory for processing
dataset.load()