BibAIFilter Benchmark Dataset & Resources for AI‑Assisted Literature Screening

Please note that the Zenodo record is currently under embargo and the files are not yet publicly accessible. It will be made publicly available once the scientific article is published. Upon your request, access permission can be evaluated and granted accordingly. This repository provides all supplementary resources used in the AI-assisted literature screening experiment involving 20 large language models. It includes the standardized prompt file, user interface demonstration video, and the human evaluation questionnaire. These materials support full reproducibility and offer a foundation for future benchmarking and comparative studies.