Interpretable Transformers via Algorithm Unrolling

Implementation and analysis of 2025 NeurIPS paper for Machine Learning on Graphs course.