Tag
This article presents a dataset and analysis pipeline for ICLR 2026 accepted papers, extracting institutional affiliations from PDF title blocks to create a clean dataset and publication-ready treemap visualizations.