MIT Press

Skip to Main Content

Table 6:

F1 of single-unit and full-hidden-state word-boundary diagnostic classifiers, trained and tested on uncontrolled running text.

	LSTM	LSTM	RNN	RNN
	single	full	single	full
English	87.7	93.0	65.6	90.5
German	86.6	91.9	70.4	85.0
Italian	85.6	92.2	71.3	91.5

	LSTM	LSTM	RNN	RNN
	single	full	single	full
English	87.7	93.0	65.6	90.5
German	86.6	91.9	70.4	85.0
Italian	85.6	92.2	71.3	91.5