Work on scaling laws has found that large language models (LMs) show
pre...
Automated reasoning with unstructured natural text is a key requirement ...
Keeping track of how states and relations of entities change as a text o...
We propose reconstruction probing, a new analysis method for contextuali...
Human linguistic capacity is often characterized by compositionality and...
Remarkable progress has been made on automated reasoning with knowledge
...
Naturally-occurring information-seeking questions often contain question...
Many Question-Answering (QA) datasets contain unanswerable questions, bu...
Natural language is characterized by compositionality: the meaning of a
...
Contextualized representation models such as ELMo (Peters et al., 2018a)...
We introduce a set of nine challenge tasks that test for the understandi...
Work on the problem of contextualized word representation -- the develop...
Distinguishing between core and non-core dependents (i.e., arguments and...