Fig. 4: Axis change for familiar faces does not depend on PR feedback to IT.
From: Temporal multiplexing of perception and memory codes in IT cortex
a, Schematic of experiment aimed at identifying the origin of memory-related signals in AM. Muscimol was injected into face patch PR and the responses of AM neurons were recorded both before and after PR inactivation using a multi-electrode probe. b, Responses of two example AM cells to screening stimuli before and after PR inactivation. c, Response profiles of the AM population to screening stimuli before and after PR inactivation. d, Time course of the similarity between preferred axes for unfamiliar–unfamiliar and unfamiliar–familiar faces (as in Fig. 2c) before and after PR inactivation. Shaded areas, s.e.m. e, Normalized firing rate changes for familiar and unfamiliar faces induced by PR inactivation (n = 85 cells). f, Familiarity decoding accuracy for screening stimuli before and after PR inactivation (n = 17 for human and monkey faces, n = 16 for objects). For objects, chi-square test χ2 (1, N = 16) = 0.58, P = 0.45. g, Face feature decoding error (mean square error) for unfamiliar (n = 1,000) and familiar (n = 36) faces before and after PR inactivation.
