Despite the centrality of the notion of representation to its explanations, neuroscience lacks a unified framework for the concepts used to characterize representation, leading to disparate use of terminology and measures associated with representation. To offer clarification, we propose a core set of conceptual dimensions that characterize representations in neuroscience. These dimensions describe relations between a neural response, features that may be represented, and downstream effects of the neural response. A neural response may be shown to be sensitive and specific to a feature, invariant to other features, and functional, which means that it is used downstream in the brain. We use information-theoretic measures to introduce these conceptual dimensions unambiguously and explain how data analysis methods such as correlational analyses, decoding and encoding models, representational similarity analysis, and tests of statistical dependence or adaptation relate to our framework. We consider several canonical examples, including the representation of orientation, numerosity, and spatial location, which illustrate how the evidence put forth in support or criticism of representational conclusions is systematized by our framework. By offering a unified conceptual framework we hope to aid the comparison and integration of results across studies and research groups and to help determine when evidence for a representational conclusion is strong.

Background: Alterations in sensory perception, a core phenotype of autism, are attributed to imbalanced integration of sensory information and prior knowledge during perceptual statistical (Bayesian) inference. This hypothesis has gained momentum in recent years, partly because it can be implemented both at the computational level, as in Bayesian perception, and at the level of canonical neural microcircuitry, as in predictive coding. However, empirical investigations have yielded conflicting results with evidence remaining limited. Critically, previous studies did not assess the independent contributions of priors and sensory uncertainty to the inference. Method: We addressed this gap by quantitatively assessing both the independent and interdependent contributions of priors and sensory uncertainty to perceptual decision-making in autistic and non-autistic individuals (N=126) during an orientation categorization task. Results: Contrary to common views, autistic individuals integrated the two Bayesian components into their decision behavior, and did so indistinguishably from non-autistic individuals. Both groups adjusted their decision criteria in a suboptimal manner. Limitations: This study focuses on explicit priors in a perceptual categorization task and high-functioning adults. Thus, although the findings provide strong evidence against a general and basic alteration in prior integration in autism, they cannot rule out more specific cases of reduced prior effect – such as due to implicit prior learning, particular level of decision making (e.g., social), and level of functioning of the autistic person. Conclusions: These results reveal intact inference for autistic individuals during perceptual decision-making, challenging the notion that Bayesian computations are fundamentally altered in autism.

We can often anticipate the precise moment when a stimulus will be relevant for our behavioral goals. Voluntary temporal attention, the prioritization of sensory information at task-relevant time points, enhances visual perception. However, the neural mechanisms of voluntary temporal attention have not been isolated from those of temporal expectation, which reflects timing predictability rather than relevance. Here we use time-resolved steady-state visual evoked responses (SSVER) to investigate how temporal attention dynamically modulates visual activity when temporal expectation is controlled. We recorded magnetoencephalography while participants directed temporal attention to one of two sequential grating targets with predictable timing. Meanwhile, a co-localized SSVER probe continuously tracked visual cortical modulations both before and after the target stimuli. We find that in the pre-target period, the SSVER gradually ramps up as the targets approach, reflecting temporal expectation. Furthermore, we find a low-frequency modulation of the SSVER, which shifts approximately half a cycle in phase according to which target is attended. In the post-target period, temporal attention to the first target transiently modulates the SSVER shortly after target onset. Thus, temporal attention dynamically modulates visual cortical responses via both periodic pre-target and transient post-target mechanisms to prioritize sensory information at precise moments. People can direct attention to specific moments that they anticipate will be relevant to their goals. Here, the authors show that voluntary temporal attention engages both periodic and transient modulations of visual cortical activity to improve perception at precise time points.

The human brain faces significant constraints in its ability to process every item in a sequence of stimuli. Voluntary temporal attention can selectively prioritize a task-relevant item over its temporal competitors to alleviate these constraints. However, it remains unclear when and where in the brain selective temporal attention modulates the visual representation of a prioritized item. Here, we manipulated temporal attention to successive stimuli in a two-target temporal cueing task, while controlling for temporal expectation with fully predictable stimulus timing. We used magnetoencephalography and time-resolved decoding to track the spatiotemporal evolution of stimulus representations in human observers. We found that temporal attention enhanced the representation of the first target around 250 ms after target onset, in a contiguous region spanning left frontal cortex and cingulate cortex. The results indicate that voluntary temporal attention recruits cortical regions beyond the ventral stream at an intermediate processing stage to amplify the representation of a target stimulus. This routing of stimulus information to anterior brain regions may provide protection from interference in visual cortex by a subsequent stimulus. Thus, voluntary temporal attention may have distinctive neural mechanisms to support specific demands of the sequential processing of stimuli.

Visual attention unfolds across space and time to prioritize a subset of incoming visual information. Distinct in key ways from spatial attention, temporal attention is a growing research area with its own conceptual and mechanistic territory. Here I review key conceptual issues, data and models in the field of visual temporal attention, with an emphasis on voluntary temporal attention. I first situate voluntary temporal attention in the broader domains of temporal attention and attentional dynamics, with the goal of organizing concepts and findings related to dynamic attention. Next, I review findings that voluntary temporal attention affects visual perception in a selective fashion - prioritizing certain time points at the expense of other time points. Selectivity is a hallmark of attention and implies a limitation in computational resources that prevents sustained maximal processing of all time points. I discuss a computational model of temporal attention that captures limited resources across time and review other models of attentional dynamics. Finally, I discuss productive future directions for the study of temporal attention. Visual temporal attention involves the prioritization of certain points in time at the expense of others. In this Review, Denison synthesizes experimental results and computational models of voluntary temporal attention and distinguishes it from related phenomena.

Attention generally enhances both visual performance and subjective appearance. Yet, at matched performance, unattended items can appear more visible than attended ones, a phenomenon called “subjective inflation.” Inflation, however, has only been narrowly tested near detection thresholds, making it unclear whether attention regularly dissociates objective and subjective aspects of perception with broad implications for everyday vision—where attention is usually unevenly distributed—and for studies of consciousness. Here, in four experiments, we tested inattentional inflation over varied stimulus and task conditions, spanning threshold to suprathreshold regimes. Using a new analytic approach to relate objective and subjective reports over full psychometric functions, we measured subjective inflation over wide ranges of matched performance. In all experiments, inattention inflated subjective stimulus visibility. But when subjective reports specified visibility of the task-relevant feature, we only found evidence for inflation at threshold. Thus, what we think we see may regularly dissociate from what we can visually discriminate. ### Competing Interest Statement The authors have declared no competing interest. Templeton World Charity Foundation, https://ror.org/00x0z1472, 0567

Vision is widely used as a model system to gain insights into how sensory inputs are processed and interpreted by the brain. Historically, careful quantification and control of visual stimuli have served as the backbone of visual neuroscience. There has been less emphasis, however, on how an observer's task influences the processing of sensory inputs. Motivated by diverse observations of task-dependent activity in the visual system, we propose a framework for thinking about tasks, their role in sensory processing, and how we might formally incorporate tasks into our models of vision.

Perceptual decision-making is often conceptualized as the process of comparing an internal decision variable to a categorical boundary or criterion. How the mind sets such a criterion has been studied from at least two perspectives. One idea is that the criterion is a fixed quantity. In work on subjective phenomenology, the notion of a fixed criterion has been proposed to explain a phenomenon called "subjective inflation"-a form of metacognitive mismatch in which observers overestimate the quality of their sensory representation in the periphery or at unattended locations. A contrasting view emerging from studies of perceptual decision-making is that the criterion adjusts to the level sensory uncertainty and is thus sensitive to variations in attention. Here, we mathematically demonstrate that previous empirical findings supporting subjective inflation are consistent with either a fixed or a flexible decision criterion. We further lay out specific task properties that are necessary to make inferences about the flexibility of the criterion: (i) a clear mapping from decision variable space to stimulus feature space and (ii) an incentive for observers to adjust their decision criterion as uncertainty changes. Recent work satisfying these requirements has demonstrated that decision criteria flexibly adjust according to uncertainty. We conclude that the fixed-criterion model of subjective inflation is poorly tenable.

All theories of perceptual decision-making postulate that external sensory information is transformed into the internal evidence that is used to guide behavior. However, the nature of this external-to-internal transformation is generally unknown. In two experiments, we examined how a particular stimulus feature–orientation–is transformed into internal evidence. Subjects judged whether Gabors were tilted clockwise or counterclockwise. The results of Experiment 1 demonstrated that increasing the stimulus tilt in fine-scale increments resulted in a linear increase in sensitivity (d’), suggesting a linear external-to-internal transformation. However, the results of Experiment 2 demonstrated that increasing the stimulus tilt in coarse-scale increments had little effect on sensitivity, suggesting a highly non-linear transformation. These results suggest that external sensory information is transformed into internal decisional evidence in complex ways. Critically, artificial neural networks (ANNs) trained on the orientation task and validated against the empirical results provided a framework for examining how sensory stimuli map onto internal evidence. The hidden-layer activations of the ANNs revealed that fine-scale increments in tilt magnitude results in increasingly greater discriminability between the stimulus categories, but the degree of discriminability does not increase further after the magnitude of stimulus tilt becomes sufficiently large. Taken together, these results begin to reveal how external sensory information is transformed into the internal evidence that is used to make decisions and suggest that ANNs could serve as a platform for understanding the mechanism underlying this critical transformation.

Introduction: Voluntary temporal attention lets us prioritize sensory information at task-relevant points in time. In vision, temporal attention improves perception at an attended moment with relative impairments before and after. Here we investigated how voluntary temporal attention affects auditory vs. visual perception to test the domain generality of selective temporal attention. Methods: We designed an auditory temporal attention experiment matched to a previous visual temporal attention experiment (Denison et al., 2017) to allow a direct comparison between the two. Two sequential auditory frequency sweeps (targets T1 and T2) with independent sweep directions (up or down) and central frequencies (800–2400 Hz) each lasted 30 ms, separated by a 250-ms stimulus onset asynchrony. A visual precue (75% validity) instructed observers to attend to T1, T2, or both (neutral precue). A visual response cue after the targets instructed observers to report the sweep direction of either T1 or T2. Sweep discrimination accuracy and reaction time were statistically compared to the previous visual results from an analogous orientation discrimination task. Results: Mean discrimination accuracy and reaction time were well-matched across the auditory and visual experiments, indicating similar levels of task difficulty. Reaction times were fastest for valid, slowest for invalid, and intermediate for neutral temporal precues in both experiments, with no difference between experiments, confirming that participants followed the precueing instructions. However, the experiments showed different effects of temporal attention on accuracy, with a stronger effect of validity on visual vs. auditory discrimination performance. Further, in the auditory task, precue validity affected performance for T1 but not for T2, whereas in the visual task, validity affected performance for both targets. Conclusion: Temporal attention affected reaction time similarly but perceptual sensitivity differently for audition and vision. The results open avenues for further research comparing the perceptual effects and timescales of auditory and visual temporal attention.

Temporal attention is the selection and prioritization of information at a specific moment. Exogenous temporal attention is the automatic, stimulus driven deployment of attention. The benefits and costs of exogenous temporal attention on performance have not been isolated. Previous experimental designs have precluded distinguishing the effects of attention and expectation about stimulus timing. Here, we manipulated exogenous temporal attention and the uncertainty of stimulus timing independently and investigated visual performance at the attended and unattended moments with different levels of temporal uncertainty. In each trial, two Gabor patches were presented consecutively with a variable stimulus onset. To drive exogenous attention and test performance at attended and unattended moments, a task-irrelevant, brief cue was presented 100 ms before target onset, and an independent response cue was presented at the end of the trial. Exogenous temporal attention slightly improved accuracy, and the effects varied with temporal uncertainty, suggesting a possible interaction of temporal attention and expectations in time.

Goal: Voluntary temporal attention, the prioritization of visual information at task-relevant points in time, improves perceptual sensitivity. However, little is known about how temporal attention affects the neural representation of visual information. Here we investigated whether and how temporal attention affects orientation representations. Methods: In two experiments, we used a two-target temporal cueing task to manipulate voluntary temporal attention and measured the effects on orientation representations using MEG decoding. On each trial, two grating targets (T1 and T2), each independently tilted around vertical or horizontal, appeared sequentially at the same location, separated by a 300-ms interval. An auditory precue (75% validity) before the targets instructed observers to attend to T1 or T2. An auditory response cue after the targets instructed observers to report the tilt (clockwise or counter-clockwise) of either T1 or T2. In trials where the precue directed temporal attention to T1, T1 was attended and T2 was unattended, and vice versa when the precue was to T2. The temporal cueing protocol was similar in the two experiments, except that in Experiment 1, the targets were presented on a gray background, whereas in Experiment 2, the targets were superimposed on 20-Hz flickering noise patches. To measure time-resolved orientation representations for each target, we decoded vertical vs. horizontal orientation across time using a linear support vector machine. Results: In both experiments, temporal attention improved behavioral accuracy, and it increased orientation decoding accuracy for T1. The improvement occurred 235-300 ms and 195-260 ms after T1 onset for Experiments 1 and 2, respectively, according to cluster permutation tests. There was no evidence for improved T2 decoding accuracy in either experiment. Conclusions: Voluntary temporal attention enhanced the orientation representation of the first target. This sensory-level change reveals a possible neural mechanism of how temporal attention could improve visual perceptual sensitivity.