nutrientdocs/grounding-benchmark

Grounding Benchmark Is a claim actually supported by the document it cites? This benchmark evaluates value grounding — whether a short hypothesis (a number, date, or fact) is entailed by a premise drawn from a real document (financial tables, filings, Wikipedia tables, prose evidence). It is a binary NLI task framed for the document-AI setting, with a deliberate focus on the case general NLI models fail: numeric and date grounding, and hard negatives that are lexically close but… See the full description on the dataset page: https://huggingface.co/datasets/nutrientdocs/grounding-benchmark.

<iframe src="https://silytics.si/embed/item/9c23560d-2555-409c-aba3-1b59196cb3bb" width="420" height="190" style="border:0" loading="lazy" title="Silytics entry"></iframe>

Silytics shows public metadata from Hugging Face and links to the original. Information may be incomplete or out of date. Last refreshed 16 October 2026. Licenses and terms of the original apply to the work itself. Not financial, legal or professional advice.