Masaumi Shimizu ¹ Shun Katakami ¹ Masato Okada ¹ Yasuko Sugase-Miyamoto ² Kazuko Hayashi ^2,3 Keiji Matsuda ² Kenichiro Miura ^4,5 Mark A. G. Eldridge ⁶ Richard Saunders ⁶ Barry J Richmond ⁶ Narihisa Matsumoto ^2*

• ¹ Graduate School of Frontier Sciences, The University of Tokyo, Kashiwa, Japan
• ² Human Informatics and Interaction Research Institute, National Institute of Advanced Industrial Science and Technology (AIST), Tsukuba, Ibaraki, Japan
• ³ Japan Society for the Promotion of Science (JSPS), Tokyo, Japan
• ⁴ Department of Pathology of Mental Diseases, National Institute of Mental Health (NIMH), Kodaira, Tōkyō, Japan
• ⁵ Section of Brain Function Information, National Institute for Physiological Sciences., Okazaki, Japan
• ⁶ Laboratory of Neuropsychology, National Institute of Mental Health, Bethesda, United States

Object categorization is a fundamental visual function, via which primates group items based on perceptual similarity. Neurons that respond to a class of complex objects, such as faces, can be found in inferior temporal cortex of macaque monkeys, comprising areas TEO and TE. The ability of monkeys to categorize cat/dog images is greatly impaired when both TE and TEO are removed, but is only modestly impaired if either region is left intact. This suggests that both TE and TEO can support object categorization. We investigated what differences exist in category information processing between areas TEO and TE. For cat and dog images, we found that category encoding performance increased during the initial phase of an image presentation, then remained stable in area TEO for the duration of the presentation in a passive fixation task. In area TE, category encoding performance continued to improve into later in the time window than in TEO. Furthermore, we found that, after cat/dog category training, TE neuronal population encode cat and dog category information more strongly than TEO population even in a fixation task (Mann-Whitney U-test, $p < 0.05$). Together, our results suggest that area TEO processes category information without changing its representation, whereas the category information representation in area TE evolves over time (both within a trial and across category training sessions), indicating that responses in TE may be influenced by not only feed-forward but top-down feedback.