Fig. 2: Trajectory k-means clustering revealed five clusters showing a conscious report effect and three patterns of interaction of exogenous attention with conscious report.

a Three neural patterns of the validity × conscious report interaction: i) three clusters (Visual, Sustained, Late accumulation) showed enhanced conscious report effect for the validly cued targets. For each cluster, the figure shows: cluster contact localization; comparison of target-locked neural activity for seen and unseen trials; comparison of interaction effect, SV: seen valid; SI: seen invalid; UV: unseen valid; UI: unseen invalid. Black horizontal bar for all ps < 0.05, Holm-Bonferroni corrected, gray shading is ±SEM across electrodes. ii) Reorienting cluster showed early sustained neural activity for invalidly cued targets. iii) a late Conscious report cluster differentiated seen from unseen targets independent of attention. Note that the Sustained cluster and the Reorienting cluster also showed significant conscious target report effects during the cueing period, before target occurrence. b Visual- and RT-modulation of target-locked neural activity. RT bins were sorted according to their mean RT from fastest to slowest, with neural activity pooled across contacts in each cluster. Magenta full curve shows mean discrimination task RT, followed by dashed magenta curve for mean subjective conscious report task RT. Neural activity in the Visual cluster synchronized with visual stimuli (cue, response screen and screen switching after discrimination task). Late accumulation cluster showed sustained neural activity until the response. Conscious report cluster exhibited sustained neural activity locked to the visual percept, but not to the report.