Github repository: [Github]
Paper: AttributionBench: How Hard is Automatic Attribution Evaluation?
Point of Contact: Yifei Li
We constructed this dataset from multiple existing data sources in a unified format, in order to create a unified and diverse testbed for evaluating advanced attribution evaluation systems. The dataset contains both in-domain training set and id-domain and out-of-domain test set.… See the full description on the dataset page:
https://huggingface.co/datasets/PaulR268/AttributionBench.
