Description: "Probing on human psycholinguistic experiments.
"we introduce a suite of diagnostics drawn from human language experiments, which allow us to ask targeted questions about information used by language
models for generating predictions in context."