Datasets:

---

EngineeringSoftware

/

PLSemanticsBench

like 1

Follow

EngineeringSoftware 11

This PR significantly enhances the dataset card for PLSemanticsBench.

Key changes include:

• Updating the license in the metadata from cc-by-2.0 to cc0-1.0 to match the GitHub repository's LICENSE file.
• Adding task_categories (text-generation, code-generation) and language (en) metadata, along with tags for programming-languages, code-reasoning, semantics, and interpreters for improved discoverability.
• Providing a comprehensive introduction to the dataset, its objectives, and the evaluation tasks from the paper abstract and the project's GitHub README.
• Linking directly to the associated paper: https://huggingface.co/papers/2510.03415
• Linking to the GitHub repository: https://github.com/EngineeringSoftware/PLSemanticsBench
• Including a "Sample Usage" section with Python code snippets to demonstrate how to run experiments using the benchmark and how to load specific dataset configurations using the datasets library. Prerequisites for running experiments, including basic installation steps, are also provided as context.
• Detailing the dataset splits in a tabular format and providing a clear data example.
• Adding the BibTeX citation for the paper.

These improvements make the dataset more informative, discoverable, and user-friendly for the community.