Fig. 5: Neural timescales dynamics reflect fine-grained abstract meaning.

A Neural timescales (\(\tau\)) estimation for three categories of lever presses: first lever press, intermediate (“stay”) lever press, and final lever press. B The change in neural timescales from resting-baseline before and after the first lever presses. Y-axis: change in neural timescales from resting-baseline (\(\varDelta \tau\)). The change in neural timescales significantly decreased from before to after the lever press in all areas (p < 0.05, one-sided Wilcoxon signed-rank test with Bonferroni correction for eight multiple comparisons). Squares: median across sessions ± s.e.m for after the event. Circles: median across sessions ± s.e.m for before the event (see supplementary Fig. 2N for sample size). C The change in neural timescales from resting-baseline before and after the final lever presses. Y-axis: change in neural timescales from resting-baseline (\(\varDelta \tau\)). The change in neural timescales significantly increased from before to after the lever press in all areas (p < 0.05, one-sided Wilcoxon signed-rank test with Bonferroni correction for eight multiple comparisons). Squares: median across sessions ± s.e.m for after the event. Circles: median across sessions ± s.e.m for before the event (see supplementary Fig. 2O for sample size). D The change in neural timescales from resting-baseline before and after the stay lever presses. The change in neural timescales did not significantly change from before to after the lever press in any area (p < 0.05, two-sided Wilcoxon signed-rank test with Bonferroni correction for 8 multiple comparisons). Squares: median across sessions ± s.e.m for after the event. Circles: median across sessions ± s.e.m for before the event (see supplementary Fig. 2P for sample size). E The change in neural timescales from resting-baseline before and after the leave lever presses. Y-axis: change in neural timescales from resting-baseline (\(\varDelta \tau\)). The change in neural timescales significantly increased from before to after the lever press in all areas (p < 0.05, one-sided Wilcoxon signed-rank test with Bonferroni correction for eight multiple comparisons). Squares: median across sessions ± s.e.m for after the event. Circles: median across sessions ± s.e.m for before the event (see supplementary Fig. 2P for sample size). OFC orbitofrontal cortex, VLPFC ventrolateral prefrontal cortex, DLPFC dorsolateral prefrontal cortex, ACC anterior cingulate cortex, FEF frontal eye fields, PM premotor cortex, SMA supplementary motor area. Source data are provided as a Source Data file.