Code for the experiments in Are Structural Concepts Universal in Transformer Language Models? Towards Interpretable Cross-Lingual Generalization (Findings of EMNLP 2023)