mteb/

SciDocsRR

SciDocsRR An MTEB dataset Massive Text Embedding Benchmark

Ranking of related scientific papers based on their title.

Task category t2t

Domains Academic, Non-fiction, Written

Reference https://allenai.org/data/scidocs

You can evaluate an embedding model on this dataset using the following code: import mteb

task = mteb.get_tasks(["SciDocsRR"]) evaluator = mteb.MTEB(task)

model = mteb.get_model(YOUR_MODEL)… See the full description on the dataset page: https://huggingface.co/datasets/mteb/SciDocsRR.