Evaluating statistical language models as pragmatic reasoners

Open in new window