SINGLE-NEURON ENCODING OF RAPIDLY LEARNED VISUAL INFORMATION RESHAPES HUMAN PERCEPTION

Humans can swiftly learn to recognize visual objects after only a few exposures. Integrating newly acquired information with existing knowledge necessitates forming enduring neuronal representations to enable future recognition. However, how rapid perceptual changes are reflected in neuronal dynamics within the human brain remains poorly understood. Here, we recorded spiking activity of neurons in medial occipital (OC) and temporal regions of the human brain (MTL) in patients performing an image recognition task involving rapid learning of degraded two-tone Mooney images. Neurons in OC and MTL modulated their firing patterns to encode rapidly learned visual information and reshape perception. Population decoding revealed that OC neurons resolved the identity of learned Mooney images at the cost of additional processing time compared to intact images, with delayed neuronal responses in the MTL. Our findings suggest that OC may not rely on feedback from MTL to support recognition following rapid perceptual learning. Instead, learning-induced dynamics observed in OC may reflect extensive recurrent processing, potentially involving top-down feedback from higher-order cortical areas, before signals reach MTL. These results highlight the need for further computation beyond bottom-up visual input representations to facilitate recognition after learning and provide spatiotemporal constraints for computational models incorporating recurrent mechanisms.