Animals actively sample their environment through actions such as whisking, sniffing, and saccadic eye movements. Computationally, sensorimotor control may be viewed as an interplay between two processes that place different demands on their neural circuits: a rapid/competitive process for promptly selecting each upcoming action, and a slow/integrative process weighing the outcomes of multiple prior actions to build a model of the environment. Using saccadic eye movements as a model system, we addressed the hypothesis that frontal and parietal cortex are computationally specialized for these two functions. Through biophysical modelling, we predicted neural signatures of the competitive and integrative processes. We localized these signals to the frontal eye fields and intra-parietal cortex, respectively, using whole-brain, high-temporal-resolution neuroimaging (MEG). This frontal/parietal specialization can be linked to the differential characteristics of cortical circuits, and thus may represent a more general organizing principle of sensorimotor function in the primate brain.

During perceptual decision-making tasks, centroparietal EEG potentials report an evidence accumulation-to-bound process that is time locked to trial onset. However, decisions in real-world environments are rarely confined to discrete trials; they instead unfold continuously, with accumulation of time-varying evidence being recency-weighted towards its immediate past. Confronted with time-varying stimuli, humans can appropriately adapt their weighting of recent evidence according to the statistics of the environment. The neural mechanisms supporting this adaptation currently remain unclear. Here, we show that humans' ability to adapt evidence weighting to different sensory environments is reflected in changes in centroparietal EEG potentials. We use a novel continuous task design to show that the Centroparietal Positivity (CPP) becomes more sensitive to fluctuations in sensory evidence when large shifts in evidence are less frequent, and is primarily sensitive to fluctuations in decision-relevant (not decision-irrelevant) sensory input. A complementary triphasic component over parietal cortex encodes the sum of recently accumulated sensory evidence, and its magnitude covaries with the duration over which different individuals integrate sensory evidence. Our findings reveal how adaptations in centroparietal responses reflect flexibility in evidence accumulation to the statistics of dynamic sensory environments.