Humans and language models diverge when predicting repeating text.

Aditya R. Vaidya, Javier Turek, Alexander Huth

Browse the full CoNLL paper archive.