1 code implementation • CoNLL (EMNLP) 2021 • Philip A. Huebner, Jon A. Willits
We conceptualize this limitation of the RNN as a failure to learn atomic internal states - states which capture information relevant to single word types without being influenced by redundant information provided by words with which they co-occur.
no code implementations • CoNLL (EMNLP) 2021 • Philip A. Huebner, Elior Sulem, Fisher Cynthia, Dan Roth
Transformer-based language models have taken the NLP world by storm.