This repository hosts code and data for our project evaluating deep learning models' capabilities for interpreting in-diagram logic using visual illusions. It encompasses our novel InDL dataset, several Jupyter notebooks showcasing our experiments, and Python scripts needed to reproduce our results.
