Magnitude estimation and stimulus discrimination tasks are affected by perceptual biases that cause the stimulus parameter to be perceived as shifted toward the mean of its distribution. These biases have been extensively studied in psychophysics and, more recently and to a lesser extent, with neural activity recordings. New computational techniques allow us to train spiking recurrent neural networks on the tasks used in the experiments. This provides us with another valuable tool with which to investigate the network mechanisms responsible for the biases and how behavior could be modeled. As an example, in this talk I will consider networks trained to discriminate the durations of temporal intervals. The trained networks presented the contraction bias, even though they were trained with a stimulus sequence without temporal correlations. The neural activity during the delay period carried information about the stimuli of the current trial and previous trials, this being one of the mechanisms that originated the contraction bias. The population activity described trajectories in a low-dimensional space and their relative locations depended on the prior distribution. The results can be modeled as an ideal observer that during the delay period sees a combination of the current and the previous stimuli. Finally, I will describe how the neural trajectories in state space encode an estimate of the interval duration. The approach could be applied to other cognitive tasks.

Medium-spiny neurons (MSNs) in the striatum are required for interval timing, or the estimation of the time over several seconds via a motor response. We and others have shown that striatal MSNs can encode the duration of temporal intervals via time-dependent ramping activity, progressive monotonic changes in firing rate preceding behaviorally salient points in time. Here, we investigated how timing-related activity within striatal ensembles changes with experience. We leveraged a rodent-optimized interval timing task in which mice ‘switch’ response ports after an amount of time has passed without reward. We report three main results. First, we found that the proportion of MSNs exhibiting time-dependent modulations of firing rate increased after 10 days of task overtraining. Second, temporal decoding by MSN ensembles increased with experience and was largely driven by time-related ramping activity. Finally, we found that time-related ramping activity generalized across both correct and error trials. These results enhance our understanding of striatal temporal processing by demonstrating that time-dependent activity within MSN ensembles evolves with experience and is dissociable from motor- and reward-related processes.