Task-dependent recurrent dynamics in visual cortex

Vector Symbolic Finite State Machines in Attractor Neural Networks

Hopfield attractor networks are robust distributed models of human memory, but they lack a general mechanism for effecting state-dependent attractor transitions in response to input. We propose construction rules such that an attractor network may implement an arbitrary finite state machine (FSM), where states and stimuli are represented by high-dimensional random vectors and all state transitions are enacted by the attractor network's dynamics. Numerical simulations show the capacity of the model, in terms of the maximum size of implementable FSM, to be linear in the size of the attractor network for dense bipolar state vectors and approximately quadratic for sparse binary state vectors. We show that the model is robust to imprecise and noisy weights, and so a prime candidate for implementation with high-density but unreliable devices. By endowing attractor networks with the ability to emulate arbitrary FSMs, we propose a plausible path by which FSMs could exist as a distributed computational primitive in biological neural networks.

Dynamics and maintenance of categorical responses in primary auditory cortex during task engagement

Grouping sets of sounds into relevant categories is an important cognitive ability that enables the association of stimuli with appropriate goal-directed behavioral responses. In perceptual tasks, the primary auditory cortex (A1) assumes a prominent role by concurrently encoding both sound sensory features and task-related variables. Here, we sought to explore the role of A1 in the initiation of sound categorization, shedding light on its involvement in this cognitive process. We trained ferrets to discriminate click trains of different rates in a Go/No-Go delayed categorization task and recorded neural activity during both active behavior and passive exposure to the same sounds. Purely categorical response components were extracted and analyzed separately from sensory responses to reveal their contributions to the overall population response throughout the trials. We found that categorical activity emerged during sound presentation in the population average and was present in both active behavioral and passive states. However, upon task engagement, categorical responses to the No-Go category became suppressed in the population code, leading to an asymmetrical representation of the Go stimuli relative to the No-Go sounds and pre-stimulus baseline. The population code underwent an abrupt change at stimulus offset, with sustained responses after the Go sounds during the delay period. Notably, the categorical responses observed during the stimulus period exhibited a significant correlation with those extracted from the delay epoch, suggesting an early involvement of A1 in stimulus categorization.

Continuous multiplexed population representations of task context in the mouse primary visual cortex

Effective task execution requires the representation of multiple task-related variables that determine how stimuli lead to correct responses. Even the primary visual cortex (V1) represents other task-related variables such as expectations, choice, and context. However, it is unclear how V1 can flexibly accommodate these variables without interfering with visual representations. We trained mice on a context-switching cross-modal decision task, where performance depends on inferring task context. We found that the context signal that emerged in V1 was behaviorally relevant as it strongly covaried with performance, independent from movement. Importantly, this signal was integrated into V1 representation by multiplexing visual and context signals into orthogonal subspaces. In addition, auditory and choice signals were also multiplexed as these signals were orthogonal to the context representation. Thus, multiplexing allows V1 to integrate visual inputs with other sensory modalities and cognitive variables to avoid interference with the visual representation while ensuring the maintenance of task-relevant variables.

Color appearance and the end of Hering's Opponent-Colors Theory

Hering's Opponent-Colors Theory has been central to understanding color appearance for 150 years. It aims to explain the phenomenology of colors with two linked propositions. First, a psychological hypothesis stipulates that any color is described necessarily and sufficiently by the extent to which it appears reddish-versus-greenish, bluish-versus-yellowish, and blackish-versus-whitish. Second, a physiological hypothesis stipulates that these perceptual mechanisms are encoded by three innate brain mechanisms. We review the evidence and conclude that neither side of the linking proposition is accurate: the theory is wrong. We sketch out an alternative, Utility-Based Coding, by which the known retinal cone-opponent mechanisms represent optimal encoding of spectral information given competing selective pressure to extract high-acuity spatial information; and phenomenological color categories represent an adaptive, efficient, output of the brain governed by behavioral demands.

Neural Mechanisms That Make Perceptual Decisions Flexible

Neural mechanisms of perceptual decision making have been extensively studied in experimental settings that mimic stable environments with repeating stimuli, fixed rules, and payoffs. In contrast, we live in an ever-changing environment and have varying goals and behavioral demands. To accommodate variability, our brain flexibly adjusts decision-making processes depending on context. Here, we review a growing body of research that explores the neural mechanisms underlying this flexibility. We highlight diverse forms of context dependency in decision making implemented through a variety of neural computations. Context-dependent neural activity is observed in a distributed network of brain structures, including posterior parietal, sensory, motor, and subcortical regions, as well as the prefrontal areas classically implicated in cognitive control. We propose that investigating the distributed network underlying flexible decisions is key to advancing our understanding and discuss a path forward for experimental and theoretical investigations.

Precise movement-based predictions in the mouse auditory cortex

Many of the sensations experienced by an organism are caused by their own actions, and accurately anticipating both the sensory features and timing of self-generated stimuli is crucial to a variety of behaviors. In the auditory cortex, neural responses to self-generated sounds exhibit frequency-specific suppression, suggesting that movement-based predictions may be implemented early in sensory processing. However, it remains unknown whether this modulation results from a behaviorally specific and temporally precise prediction, nor is it known whether corresponding expectation signals are present locally in the auditory cortex. To address these questions, we trained mice to expect the precise acoustic outcome of a forelimb movement using a closed-loop sound-generating lever. Dense neuronal recordings in the auditory cortex revealed suppression of responses to self-generated sounds that was specific to the expected acoustic features, to a precise position within the movement, and to the movement that was coupled to sound during training. Prediction-based suppression was concentrated in L2/3 and L5, where deviations from expectation also recruited a population of prediction-error neurons that was otherwise unresponsive. Recording in the absence of sound revealed abundant movement signals in deep layers that were biased toward neurons tuned to the expected sound, as well as expectation signals that were present throughout the cortex and peaked at the time of expected auditory feedback. Together, these findings identify distinct populations of auditory cortical neurons with movement, expectation, and error signals consistent with a learned internal model linking an action to its specific acoustic outcome.

Cortical feedback loops bind distributed representations of working memory

Working memory—the brain’s ability to internalize information and use it flexibly to guide behaviour—is an essential component of cognition. Although activity related to working memory has been observed in several brain regions^1–3, how neural populations actually represent working memory^4–7 and the mechanisms by which this activity is maintained^8–12 remain unclear^13–15. Here we describe the neural implementation of visual working memory in mice alternating between a delayed non-match-to-sample task and a simple discrimination task that does not require working memory but has identical stimulus, movement and reward statistics. Transient optogenetic inactivations revealed that distributed areas of the neocortex were required selectively for the maintenance of working memory. Population activity in visual area AM and premotor area M2 during the delay period was dominated by orderly low-dimensional dynamics^16,17 that were, however, independent of working memory. Instead, working memory representations were embedded in high-dimensional population activity, present in both cortical areas, persisted throughout the inter-stimulus delay period, and predicted behavioural responses during the working memory task. To test whether the distributed nature of working memory was dependent on reciprocal interactions between cortical regions^18–20, we silenced one cortical area (AM or M2) while recording the feedback it received from the other. Transient inactivation of either area led to the selective disruption of inter-areal communication of working memory. Therefore, reciprocally interconnected cortical areas maintain bound high-dimensional representations of working memory.

Experiments in mice alternating between a visual working memory task and a task that is independent of working memory provide insight into the neural representation of working memory and the distributed nature of its maintenance.

Neural signature of the perceptual decision in the neural population responses of the inferior temporal cortex

Rapid categorization of visual objects is critical for comprehending our complex visual world. The role of individual cortical neurons and neural populations in categorizing visual objects during passive vision has previously been studied. However, it is unclear whether and how perceptually guided behaviors affect the encoding of stimulus categories by neural population activity in the higher visual cortex. Here we studied the activity of the inferior temporal (IT) cortical neurons in macaque monkeys during both passive viewing and categorization of ambiguous body and object images. We found enhanced category information in the IT neural population activity during the correct, but not wrong, trials of the categorization task compared to the passive task. This encoding enhancement was task difficulty dependent with progressively larger values in trials with more ambiguous stimuli. Enhancement of IT neural population information for behaviorally relevant stimulus features suggests IT neural networks' involvement in perceptual decision-making behavior.

Task-induced neural covariability as a signature of approximate Bayesian learning and inference

Perception is often characterized computationally as an inference process in which uncertain or ambiguous sensory inputs are combined with prior expectations. Although behavioral studies have shown that observers can change their prior expectations in the context of a task, robust neural signatures of task-specific priors have been elusive. Here, we analytically derive such signatures under the general assumption that the responses of sensory neurons encode posterior beliefs that combine sensory inputs with task-specific expectations. Specifically, we derive predictions for the task-dependence of correlated neural variability and decision-related signals in sensory neurons. The qualitative aspects of our results are parameter-free and specific to the statistics of each task. The predictions for correlated variability also differ from predictions of classic feedforward models of sensory processing and are therefore a strong test of theories of hierarchical Bayesian inference in the brain. Importantly, we find that Bayesian learning predicts an increase in so-called “differential correlations” as the observer’s internal model learns the stimulus distribution, and the observer’s behavioral performance improves. This stands in contrast to classic feedforward encoding/decoding models of sensory processing, since such correlations are fundamentally information-limiting. We find support for our predictions in data from existing neurophysiological studies across a variety of tasks and brain areas. Finally, we show in simulation how measurements of sensory neural responses can reveal information about a subject’s internal beliefs about the task. Taken together, our results reinterpret task-dependent sources of neural covariability as signatures of Bayesian inference and provide new insights into their cause and their function.

Perceptual decision-making has classically been studied in the context of feedforward encoding/ decoding models. Here, we derive predictions for the responses of sensory neurons under the assumption that the brain performs hierarchical Bayesian inference, including feedback signals that communicate task-specific prior expectations. Interestingly, those predictions stand in contrast to some of the conclusions drawn in the classic framework. In particular, we find that Bayesian learning predicts the increase of a type of correlated variability called “differential correlations” over the course of learning. Differential correlations limit information, and hence are seen as harmful in feedforward models. Since our results are also specific to the statistics of a given task, and since they hold under a wide class of theories about how Bayesian probabilities may be represented by neural responses, they constitute a strong test of the Bayesian Brain hypothesis. Our results can explain the task-dependence of correlated variability in prior studies and suggest a reason why these kinds of correlations are surprisingly common in empirical data. Interpreted in a probabilistic framework, correlated variability provides a window into an observer’s task-related beliefs.

Confidence-Controlled Hebbian Learning Efficiently Extracts Category Membership From Stimuli Encoded in View of a Categorization Task

In experiments on perceptual decision making, individuals learn a categorization task through trial-and-error protocols. We explore the capacity of a decision-making attractor network to learn a categorization task through reward-based, Hebbian-type modifications of the weights incoming from the stimulus encoding layer. For the latter, we assume a standard layer of a large number of stimulus-specific neurons. Within the general framework of Hebbian learning, we have hypothesized that the learning rate is modulated by the reward at each trial. Surprisingly, we find that when the coding layer has been optimized in view of the categorization task, such reward-modulated Hebbian learning (RMHL) fails to extract efficiently the category membership. In previous work, we showed that the attractor neural networks' nonlinear dynamics accounts for behavioral confidence in sequences of decision trials. Taking advantage of these findings, we propose that learning is controlled by confidence, as computed from the neural activity of the decision-making attractor network. Here we show that this confidence-controlled, reward-based Hebbian learning efficiently extracts categorical information from the optimized coding layer. The proposed learning rule is local and, in contrast to RMHL, does not require storing the average rewards obtained on previous trials. In addition, we find that the confidence-controlled learning rule achieves near-optimal performance. In accordance with this result, we show that the learning rule approximates a gradient descent method on a maximizing reward cost function.

A confirmation bias in perceptual decision-making due to hierarchical approximate inference

Making good decisions requires updating beliefs according to new evidence. This is a dynamical process that is prone to biases: in some cases, beliefs become entrenched and resistant to new evidence (leading to primacy effects), while in other cases, beliefs fade over time and rely primarily on later evidence (leading to recency effects). How and why either type of bias dominates in a given context is an important open question. Here, we study this question in classic perceptual decision-making tasks, where, puzzlingly, previous empirical studies differ in the kinds of biases they observe, ranging from primacy to recency, despite seemingly equivalent tasks. We present a new model, based on hierarchical approximate inference and derived from normative principles, that not only explains both primacy and recency effects in existing studies, but also predicts how the type of bias should depend on the statistics of stimuli in a given task. We verify this prediction in a novel visual discrimination task with human observers, finding that each observer's temporal bias changed as the result of changing the key stimulus statistics identified by our model. The key dynamic that leads to a primacy bias in our model is an overweighting of new sensory information that agrees with the observer's existing belief-a type of 'confirmation bias'. By fitting an extended drift-diffusion model to our data we rule out an alternative explanation for primacy effects due to bounded integration. Taken together, our results resolve a major discrepancy among existing perceptual decision-making studies, and suggest that a key source of bias in human decision-making is approximate hierarchical inference.

Linear Integration of Sensory Evidence over Space and Time Underlies Face Categorization

Visual object recognition relies on elaborate sensory processes that transform retinal inputs to object representations, but it also requires decision-making processes that read out object representations and function over prolonged time scales. The computational properties of these decision-making processes remain underexplored for object recognition. Here, we study these computations by developing a stochastic multifeature face categorization task. Using quantitative models and tight control of spatiotemporal visual information, we demonstrate that human subjects (five males, eight females) categorize faces through an integration process that first linearly adds the evidence conferred by task-relevant features over space to create aggregated momentary evidence and then linearly integrates it over time with minimum information loss. Discrimination of stimuli along different category boundaries (e.g., identity or expression of a face) is implemented by adjusting feature weights of spatial integration. This linear but flexible integration process over space and time bridges past studies on simple perceptual decisions to complex object recognition behavior.SIGNIFICANCE STATEMENT Although simple perceptual decision-making such as discrimination of random dot motion has been successfully explained as accumulation of sensory evidence, we lack rigorous experimental paradigms to study the mechanisms underlying complex perceptual decision-making such as discrimination of naturalistic faces. We develop a stochastic multifeature face categorization task as a systematic approach to quantify the properties and potential limitations of the decision-making processes during object recognition. We show that human face categorization could be modeled as a linear integration of sensory evidence over space and time. Our framework to study object recognition as a spatiotemporal integration process is broadly applicable to other object categories and bridges past studies of object recognition and perceptual decision-making.

Choices change the temporal weighting of decision evidence

Many decisions result from the accumulation of decision-relevant information (evidence) over time. Even when maximizing decision accuracy requires weighting all the evidence equally, decision-makers often give stronger weight to evidence occurring early or late in the evidence stream. Here, we show changes in such temporal biases within participants as a function of intermittent judgments about parts of the evidence stream. Human participants performed a decision task that required a continuous estimation of the mean evidence at the end of the stream. The evidence was either perceptual (noisy random dot motion) or symbolic (variable sequences of numbers). Participants also reported a categorical judgment of the preceding evidence half-way through the stream in one condition or executed an evidence-independent motor response in another condition. The relative impact of early versus late evidence on the final estimation flipped between these two conditions. In particular, participants' sensitivity to late evidence after the intermittent judgment, but not the simple motor response, was decreased. Both the intermittent response as well as the final estimation reports were accompanied by nonluminance-mediated increases of pupil diameter. These pupil dilations were bigger during intermittent judgments than simple motor responses and bigger during estimation when the late evidence was consistent than inconsistent with the initial judgment. In sum, decisions activate pupil-linked arousal systems and alter the temporal weighting of decision evidence. Our results are consistent with the idea that categorical choices in the face of uncertainty induce a change in the state of the neural circuits underlying decision-making.NEW & NOTEWORTHY The psychology and neuroscience of decision-making have extensively studied the accumulation of decision-relevant information toward a categorical choice. Much fewer studies have assessed the impact of a choice on the processing of subsequent information. Here, we show that intermittent choices during a protracted stream of input reduce the sensitivity to subsequent decision information and transiently boost arousal. Choices might trigger a state change in the neural machinery for decision-making.

Decoding and encoding (de)mixed population responses

A central tenet of neuroscience is that the brain works through large populations of interacting neurons. With recent advances in recording techniques, the inner working of these populations has come into full view. Analyzing the resulting large-scale data sets is challenging because of the often complex and 'mixed' dependency of neural activities on experimental parameters, such as stimuli, decisions, or motor responses. Here we review recent insights gained from analyzing these data with dimensionality reduction methods that 'demix' these dependencies. We demonstrate that the mappings from (carefully chosen) experimental parameters to population activities appear to be typical and stable across tasks, brain areas, and animals, and are often identifiable by linear methods. By considering when and why dimensionality reduction and demixing work well, we argue for a view of population coding in which populations represent (demixed) latent signals, corresponding to stimuli, decisions, motor responses, and so on. These latent signals are encoded into neural population activity via non-linear mappings and decoded via linear readouts. We explain how such a scheme can facilitate the propagation of information across cortical areas, and we review neural network architectures that can reproduce the encoding and decoding of latent signals in population activities. These architectures promise a link from the biophysics of single neurons to the activities of neural populations.