Adaptive behaviour crucially depends on flexible decision-making, which in mammals relies on the frontal cortex, specifically the orbitofrontal cortex (OFC) 1−9 . How OFC encodes decision variables and instructs sensory areas to guide adaptive behaviour are key open questions. Here we developed a reversal learning task for head-fixed mice, monitored the activity of neurons of the lateral OFC using twophoton calcium imaging and investigated how OFC dynamically interacts with primary somatosensory cortex (S1). Mice learned to discriminate 'go' from 'no-go' tactile stimuli 10,11 and adapt their behaviour upon reversal of stimulus-reward contingency ('rule switch'). Imaging individual neurons longitudinally across all behavioural phases revealed a distinct engagement of S1 and lateral OFC, with S1 neural activity reflecting initial task learning, whereas lateral OFC neurons responded saliently and transiently to the rule switch. We identified direct long-range projections from lateral OFC to S1 that can feed this activity back to S1 as value prediction error. This top-down signal updated sensory representations in S1 by functionally remapping responses in a subpopulation of neurons that was sensitive to reward history. Functional remapping crucially depended on top-down feedback as chemogenetic silencing of lateral OFC neurons disrupted reversal learning, as well as plasticity in S1. The dynamic interaction of lateral OFC with sensory cortex thus implements computations critical for value prediction that are history dependent and error based, providing plasticity essential for flexible decision-making.

Flexible decision-making is crucial for adaptive behaviour. Such behaviour in mammals largely relies on the frontal cortex, and specifically, the orbitofrontal cortex (OFC). How OFC neurons encode decision variables and instruct sensory areas to guide adaptive behaviour is a key open question. Here we developed a reversal learning task for head-fixed mice together with twophoton calcium imaging to monitor the activity of lateral OFC neuronal populations and investigated their dynamic interaction with primary somatosensory cortex (S1). Mice trained on this task learned to discriminate go/no-go tactile stimuli and adapt their behaviour upon changes in stimulusreward contingencies ('rule-switch'). Longitudinal imaging at cellular resolution across weeks during all behavioural phases revealed a distinct engagement of S1 and lateral OFC neurons: S1 neural activity reflected task learning-related responses, while neurons in the lateral OFC saliently and transiently responded to the rule-switch. A subset of OFC neurons conveyed a value prediction error signal via feedback projections to S1, as direct anatomical long-range projections were revealed by retrograde tracing combined with whole-brain light-sheet microscopy. Top-down signals implemented an update of sensory representations and functionally reconfigured a small subpopulation of S1 neurons that were differentially modulated by reward-history. Functional remapping of these neurons crucially depended on top-down inputs, as chemogenetic silencing of lateral OFC neurons disrupted reversal learning and impaired plastic changes in these outcome-sensitive S1 neurons. Our results reveal the presence of long-range cortical interactions between cellular ensembles in higher and lower-order brain areas specifically recruited during context-dependent learning and taskswitching. Such interactions crucially implement history-dependent rewardvalue computations and error heuristics, which, in turn, help guide adaptive behaviour. analysed by C.L.). A.B and G.P. analysed data. F.F.V developed light-sheet microscope and imaged cleared brains together with A.B. A.B. and F.H. wrote the manuscript with comments from all authors.