We present GENTLE, a new mixed-genre English challenge corpus totaling 17K tokens and consisting of 8 unusual text types for out-of-domain evaluation: dictionary entries, esports commentaries, legal documents, medical notes, poetry, mathematical proofs, syllabuses, and threat letters. GENTLE is manually annotated for a variety of popular NLP tasks, including syntactic dependency parsing, entity recognition, coreference resolution, and discourse parsing. We evaluate state-of-the-art NLP systems on GENTLE and find severe degradation for at least some genres in their performance on all tasks, which indicates GENTLE’s utility as an evaluation dataset for NLP systems.

The current study quantitatively (and qualitatively for an illustrative purpose) analyzes BERT’s layer-wise masked word prediction on an English corpus, and finds that (1) the layerwise localization of linguistic knowledge primarily shown in probing studies is replicated in a behavior-based design and (2) that syntactic and semantic information is encoded at different layers for words of different syntactic categories. Hypothesizing that the above results are correlated with the number of likely potential candidates of the masked word prediction, we also investigate how the results differ for tokens within multiword expressions.