Replication Package: On the Generalization Capabilities of LLMs for Reverse Engineering Sequence Diagrams

This archive contains the replication package for our paper On the Generalization Capabilities of LLMs for Reverse Engineering Sequence Diagrams accepted at 7th Workshop on Artificial Intelligence and Model-driven Engineering (MDE Intelligence). The paper will be published with IEEE.This dataset comprises the zipped source code of the Github Release including all software to conduct the experiment the Python sequence diagrams generated by our finetuned model: seq_dataset_filtered_eval.zip The curated training and evaluation dataset as well as the finetuned model are separately published on Zenodo (cf, supplementary data) as also described in the README of the source code repository.