Fig. 8: Fine-grained temporal sequence of speech encoding during production and comprehension. | Nature Human Behaviour

Fig. 8: Fine-grained temporal sequence of speech encoding during production and comprehension.

From: A unified acoustic-to-speech-to-language embedding space captures the neural basis of natural language processing in everyday conversations

Fig. 8

a, Encoding models for encoder units 1–20 time locked to word onset (corresponding to a temporal segment of 20–400 ms after word onset) during production. The encoding performance exhibits two peaks (one before and one after word onset). Data are presented as mean ± s.e. across electrodes. b, Encoding models for encoder units 1–20 time locked to word onset during comprehension. The encoding performance peaks mainly after word onset. Data are presented as mean ± s.e. across electrodes. c, Coloured squares correspond to peaks encoding during production before word onset, and round dots after word onset. The model was found to be significant, the P value is two-sided. d, Coloured dots correspond to peaks encoding during comprehension after word onset. Data are presented as mean ± s.e.m. across electrodes. The model was found to be significant, the P value is two-sided.

Back to article page