Abstract
Making perceptual decisions to classify unknown sensory stimuli into behaviorally relevant categories, known as category learning, is a fundamental cognitive function1-4. The posterior parietal cortex (PPC), although has been intensively studied for its role in perceptual and categorical decision-making5-10, as well as in many other cognitive functions11-17, its functional significance remains controversial due to lack of causal effects on choice behaviors10,18-21. Here we show that PPC neurons exhibit characteristic dynamics of category learning, and are causally involved specifically in the decision process of categorizing new sensory stimuli, which is attributable to the PPC to sensory cortex feedback circuits. We trained mice to perform an auditory category learning task, where mice made categorical decisions classifying new tone stimuli into learned categories. Using in vivo two-photon imaging tracking the activity of the same population of PPC neurons over learning, we found that PPC neurons show dynamic representations adaptively classifying new stimuli while maintaining stable representation for learned categories. Inactivation of PPC using pharmacology or optogenetics led to significant impairment in categorical decisions for new stimuli but not for well-learned stimuli. In addition, PPC inactivation led to a stronger bias of choices by previous trials, revealing an additional causal contribution of PPC counterbalancing recent history bias. Circuit-level perturbations using chemogenetics to silence the PPC to dorsal auditory cortex projections also impaired categorical decisions during learning. Our results reveal a dynamic and causal role of PPC circuits in decision-making when categorizing new sensory stimuli, which not only provide a potential reconciliation to the widespread controversies, but also provide a circuit-level mechanism for the key cognitive function of perceptual classification of new sensory information.