Auditory cortex reflects goal-directed movement but is not necessary for behavioral adaptation in sound-cued reward tracking

Abstract

Recent evidence suggests that sensory cortex represents nonsensory variables such as reward expectation, but the relevance of these representations for behavior are not well understood. We show that rat auditory cortex (AC) is modulated during movement and reward anticipation in a sound-cued reward tracking task, whereas AC inactivation only impaired discrimination without affecting reward tracking, consistent with a predominantly sensory role of AC.

Integration of sensory evidence and reward expectation in mouse perceptual decision-making task with various sensory uncertainties

“…In humans, choice biases are often smaller than the optimal ( Maddox, 2002 ). This is consistent with the findings in rodents including our study ( Figure 2 ) ( Stoilova et al., 2020 ). In contrast, monkeys had larger choice biases than the optimal ( Feng et al., 2009 ; Teichert and Ferrera, 2010 ).…”

“…Adding a parameter in the simple psychometric function always increased the log likelihood; the likelihood ratio test investigated whether the additional parameter significantly improved the fitting (Δ log likelihood in Figure 2 D, left) ( Daw, 2011 ; Maddox, 2002 ). We verified that the psychometric function adding the block-dependent decision threshold (block threshold) significantly improved the fitting than the simple psychometric function (likelihood ratio test, p = 1.8 × 10 -10 ) ( Figure 2 D, left), indicating that the asymmetric reward blocks biased the choices as reported in previous studies ( Daw, 2011 ; Lak et al., 2020a ; Nomoto et al., 2010 ; Stoilova et al., 2020 ; Stüttgen et al., 2011 ). We then found that the psychometric function with block-dependent thresholds and duration-dependent sensitivities fit the choices (likelihood ratio test, p = 0.0073), but the parameters for duration-dependent thresholds (p = 0.048), block-dependent sensitivities (p = 0.041), or lapse rate (p = 0.35) did not significantly improve the fitting after Bonferroni correction ( Figure 2 D, right).…”

“…This gradual updating of choice bias was captured by our RL model ( Figure 4 ). Previous studies also found that prior knowledge of stimulus probability and reward amount was updated trial by trial in an RL rule to bias choices ( Bell et al., 2016 ; Lak et al., 2020a ; Stoilova et al., 2020 ).…”

“…When sensory inputs are uncertain, Bayesian theory shows that integration of sensory stimuli with prior knowledge of stimulus probability and action outcomes is essential to optimize behavior ( Dayan and Daw, 2008 ; Gold and Ding, 2013 ; Körding et al., 2004 ), as the prediction of serve placement (stimulus probability) and the success rate of forehands and backhands (action outcome) are the keys to returning the first serve. Previous studies have tested the integration of sensory inputs and prior knowledge with a sensory discrimination task by biasing the reward outcomes or stimulus probabilities in humans ( Maddox, 2002 ), monkeys ( Feng et al., 2009 ; Gold and Shadlen, 2007 ; Hanks et al., 2011 ; Nomoto et al., 2010 ; Platt and Glimcher, 1999 ; Teichert and Ferrera, 2010 ), pigeons ( McCarthy and Davison, 1979 ; Stüttgen et al., 2011 ), and rodents ( Aguillon-Rodriguez et al., 2021 ; Lak et al., 2020a ; Stoilova et al., 2020 ). In these studies, subjects tended to select choices associated with a large reward amount or high stimulus probability, indicating that animals integrate the sensory stimulus and the prior.…”

“…Here, we developed a single interval forced choice (SIFC) task ( Stoilova et al., 2020 ) for head-fixed mice based on a previous study ( Marbach and Zador, 2017 ), in which mice discriminated a tone frequency to choose between a spout on the left and another on the right. The SIFC task is also known as yes-no task ( McCarthy and Davison, 1979 ; Yeshurun et al., 2008 ).…”