Automatic Generation of Safety-compliant Linear Temporal Logic via Large Language Model: A Self-supervised Framework - QEST + FORMATS 2025 Artifact

Relation to paper The Evaluation section of our paper contains six experiments (Table 6): nl2spec AutoSafeLTL AutoSafeLTL-B nl2spec+B AutoSafeLTL-C AutoSafeLTL-D The six experiments can be reproduced by utilizing the codes and datasets in folder 'Artifacts' and following the instruction. Aritfact structure This artifact consists of the following files and folders: File README.pdf: a thorough instruction for reproducing the experiments, including Environment Setting, Replicating the Results in Our Paper, and Customization of the Source Code to Your Own Example and Additional Information. File LICENSE.odf: the lience for download and evaluating this artifact. Folder 'Artifacts/Code': includes all the source code files needed for reproducing the experiments. Folder 'Artifacts/Dataset': includes all the datasets used in the code files. File 'rabit250-new.zip': an extend version of rabit tool utlized in the code files.