Description: "Our experiments are based on the “edge probing” approach,  which aims to measure how well information about linguistic structure can be extracted from a pre-trained encoder."

"We use eight labeling tasks from the edge
probing suite: part-of-speech (POS), constituents
(Consts.), dependencies (Deps.), entities, semantic role labeling (SRL), coreference (Coref.), semantic proto-roles (SPR; Reisinger et al., 2015), and relation classification (SemEval). "