timo-pierre-schrader/

QUITE

	Dataset Card for QUITE





	
	Dataset Description

QUITE (Quantifying Uncertainty in natural language Text) is an entirely new benchmark that allows for assessing the capabilities of neural language model-based systems w.r.t. to Bayesian reasoning on a large set of input text that describes probabilistic relationships in natural language text. For example, take the following statement from QUITE:

If Plcg is in a high state, PIP3 appears in a low state in 42% of all cases, in an… See the full description on the dataset page: https://huggingface.co/datasets/timo-pierre-schrader/QUITE.