jiawennnn/

PASS-benchmark

Data used in the PASS study: bioRxiv/medRxiv preprints paired with the journal they were later published in, the candidate-journal target sets, the ablation cohort, and journal-level SCImago metrics. Ground-truth publication status was recovered from the bioRxiv/medRxiv APIs + CrossRef.

	Files




	


	Benchmark data (2,184 preprints) — evaluation_papers.csv

The papers used for evaluation: newest version preprints whose… See the full description on the dataset page: https://huggingface.co/datasets/jiawennnn/PASS-benchmark.