Correlations and Neuronal Population Information.

Brain function involves the activity of neuronal populations. Much recent effort has been devoted to measuring the activity of neuronal populations in different parts of the brain under various experimental conditions. Population activity patterns contain rich structure, yet many studies have focused on measuring pairwise relationships between members of a larger population-termed noise correlations. Here we review recent progress in understanding how these correlations affect population information, how information should be quantified, and what mechanisms may give rise to correlations. As population coding theory has improved, it has made clear that some forms of correlation are more important for information than others. We argue that this is a critical lesson for those interested in neuronal population responses more generally: Descriptions of population responses should be motivated by and linked to well-specified function. Within this context, we offer suggestions of where current theoretical frameworks fall short.

Investigating the representation of uncertainty in neuronal circuits.

Skilled behavior often displays signatures of Bayesian inference. In order for the brain to implement the required computations, neuronal activity must carry accurate information about the uncertainty of sensory inputs. Two major approaches have been proposed to study neuronal representations of uncertainty. The first one, the Bayesian decoding approach, aims primarily at decoding the posterior probability distribution of the stimulus from population activity using Bayes' rule, and indirectly yields uncertainty estimates as a by-product. The second one, which we call the correlational approach, searches for specific features of neuronal activity (such as tuning-curve width and maximum firing-rate) which correlate with uncertainty. To compare these two approaches, we derived a new normative model of sound source localization by Interaural Time Difference (ITD), that reproduces a wealth of behavioral and neural observations. We found that several features of neuronal activity correlated with uncertainty on average, but none provided an accurate estimate of uncertainty on a trial-by-trial basis, indicating that the correlational approach may not reliably identify which aspects of neuronal responses represent uncertainty. In contrast, the Bayesian decoding approach reveals that the activity pattern of the entire population was required to reconstruct the trial-to-trial posterior distribution with Bayes' rule. These results suggest that uncertainty is unlikely to be represented in a single feature of neuronal activity, and highlight the importance of using a Bayesian decoding approach when exploring the neural basis of uncertainty.

Learning optimal decisions with confidence.

Diffusion decision models (DDMs) are immensely successful models for decision making under uncertainty and time pressure. In the context of perceptual decision making, these models typically start with two input units, organized in a neuron-antineuron pair. In contrast, in the brain, sensory inputs are encoded through the activity of large neuronal populations. Moreover, while DDMs are wired by hand, the nervous system must learn the weights of the network through trial and error. There is currently no normative theory of learning in DDMs and therefore no theory of how decision makers could learn to make optimal decisions in this context. Here, we derive such a rule for learning a near-optimal linear combination of DDM inputs based on trial-by-trial feedback. The rule is Bayesian in the sense that it learns not only the mean of the weights but also the uncertainty around this mean in the form of a covariance matrix. In this rule, the rate of learning is proportional (respectively, inversely proportional) to confidence for incorrect (respectively, correct) decisions. Furthermore, we show that, in volatile environments, the rule predicts a bias toward repeating the same choice after correct decisions, with a bias strength that is modulated by the previous choice's difficulty. Finally, we extend our learning rule to cases for which one of the choices is more likely a priori, which provides insights into how such biases modulate the mechanisms leading to optimal decisions in diffusion models.

Brain plasticity through the life span: learning to learn and action video games.

The ability of the human brain to learn is exceptional. Yet, learning is typically quite specific to the exact task used during training, a limiting factor for practical applications such as rehabilitation, workforce training, or education. The possibility of identifying training regimens that have a broad enough impact to transfer to a variety of tasks is thus highly appealing. This work reviews how complex training environments such as action video game play may actually foster brain plasticity and learning. This enhanced learning capacity, termed learning to learn, is considered in light of its computational requirements and putative neural mechanisms.

Inferring decoding strategies for multiple correlated neural populations.

Studies of neuron-behaviour correlation and causal manipulation have long been used separately to understand the neural basis of perception. Yet these approaches sometimes lead to drastically conflicting conclusions about the functional role of brain areas. Theories that focus only on choice-related neuronal activity cannot reconcile those findings without additional experiments involving large-scale recordings to measure interneuronal correlations. By expanding current theories of neural coding and incorporating results from inactivation experiments, we demonstrate here that it is possible to infer decoding weights of different brain areas at a coarse scale without precise knowledge of the correlation structure. We apply this technique to neural data collected from two different cortical areas in macaque monkeys trained to perform a heading discrimination task. We identify two opposing decoding schemes, each consistent with data depending on the nature of correlated noise. Our theory makes specific testable predictions to distinguish these scenarios experimentally without requiring measurement of the underlying noise correlations.

Robust information propagation through noisy neural circuits.

Sensory neurons give highly variable responses to stimulation, which can limit the amount of stimulus information available to downstream circuits. Much work has investigated the factors that affect the amount of information encoded in these population responses, leading to insights about the role of covariability among neurons, tuning curve shape, etc. However, the informativeness of neural responses is not the only relevant feature of population codes; of potentially equal importance is how robustly that information propagates to downstream structures. For instance, to quantify the retina's performance, one must consider not only the informativeness of the optic nerve responses, but also the amount of information that survives the spike-generating nonlinearity and noise corruption in the next stage of processing, the lateral geniculate nucleus. Our study identifies the set of covariance structures for the upstream cells that optimize the ability of information to propagate through noisy, nonlinear circuits. Within this optimal family are covariances with "differential correlations", which are known to reduce the information encoded in neural population activities. Thus, covariance structures that maximize information in neural population codes, and those that maximize the ability of this information to propagate, can be very different. Moreover, redundancy is neither necessary nor sufficient to make population codes robust against corruption by noise: redundant codes can be very fragile, and synergistic codes can-in some cases-optimize robustness against noise.

Origin of information-limiting noise correlations.

The ability to discriminate between similar sensory stimuli relies on the amount of information encoded in sensory neuronal populations. Such information can be substantially reduced by correlated trial-to-trial variability. Noise correlations have been measured across a wide range of areas in the brain, but their origin is still far from clear. Here we show analytically and with simulations that optimal computation on inputs with limited information creates patterns of noise correlations that account for a broad range of experimental observations while at same time causing information to saturate in large neural populations. With the example of a network of V1 neurons extracting orientation from a noisy image, we illustrate to our knowledge the first generative model of noise correlations that is consistent both with neurophysiology and with behavioral thresholds, without invoking suboptimal encoding or decoding or internal sources of variability such as stochastic network dynamics or cortical state fluctuations. We further show that when information is limited at the input, both suboptimal connectivity and internal fluctuations could similarly reduce the asymptotic information, but they have qualitatively different effects on correlations leading to specific experimental predictions. Our study indicates that noise at the sensory periphery could have a major effect on cortical representations in widely studied discrimination tasks. It also provides an analytical framework to understand the functional relevance of different sources of experimentally measured correlations.

Action video game play facilitates the development of better perceptual templates.

The field of perceptual learning has identified changes in perceptual templates as a powerful mechanism mediating the learning of statistical regularities in our environment. By measuring threshold-vs.-contrast curves using an orientation identification task under varying levels of external noise, the perceptual template model (PTM) allows one to disentangle various sources of signal-to-noise changes that can alter performance. We use the PTM approach to elucidate the mechanism that underlies the wide range of improvements noted after action video game play. We show that action video game players make use of improved perceptual templates compared with nonvideo game players, and we confirm a causal role for action video game play in inducing such improvements through a 50-h training study. Then, by adapting a recent neural model to this task, we demonstrate how such improved perceptual templates can arise from reweighting the connectivity between visual areas. Finally, we establish that action gamers do not enter the perceptual task with improved perceptual templates. Instead, although performance in action gamers is initially indistinguishable from that of nongamers, action gamers more rapidly learn the proper template as they experience the task. Taken together, our results establish for the first time to our knowledge the development of enhanced perceptual templates following action game play. Because such an improvement can facilitate the inference of the proper generative model for the task at hand, unlike perceptual learning that is quite specific, it thus elucidates a general learning mechanism that can account for the various behavioral benefits noted after action game play.

Measuring Fisher information accurately in correlated neural populations.

Neural responses are known to be variable. In order to understand how this neural variability constrains behavioral performance, we need to be able to measure the reliability with which a sensory stimulus is encoded in a given population. However, such measures are challenging for two reasons: First, they must take into account noise correlations which can have a large influence on reliability. Second, they need to be as efficient as possible, since the number of trials available in a set of neural recording is usually limited by experimental constraints. Traditionally, cross-validated decoding has been used as a reliability measure, but it only provides a lower bound on reliability and underestimates reliability substantially in small datasets. We show that, if the number of trials per condition is larger than the number of neurons, there is an alternative, direct estimate of reliability which consistently leads to smaller errors and is much faster to compute. The superior performance of the direct estimator is evident both for simulated data and for neuronal population recordings from macaque primary visual cortex. Furthermore we propose generalizations of the direct estimator which measure changes in stimulus encoding across conditions and the impact of correlations on encoding and decoding, typically denoted by Ishuffle and Idiag respectively.

Multisensory decisions provide support for probabilistic number representations.

A large body of evidence suggests that an approximate number sense allows humans to estimate numerosity in sensory scenes. This ability is widely observed in humans, including those without formal mathematical training. Despite this, many outstanding questions remain about the nature of the numerosity representation in the brain. Specifically, it is not known whether approximate numbers are represented as scalar estimates of numerosity or, alternatively, as probability distributions over numerosity. In the present study, we used a multisensory decision task to distinguish these possibilities. We trained human subjects to decide whether a test stimulus had a larger or smaller numerosity compared with a fixed reference. Depending on the trial, the numerosity was presented as either a sequence of visual flashes or a sequence of auditory tones, or both. To test for a probabilistic representation, we varied the reliability of the stimulus by adding noise to the visual stimuli. In accordance with a probabilistic representation, we observed a significant improvement in multisensory compared with unisensory trials. Furthermore, a trial-by-trial analysis revealed that although individual subjects showed strategic differences in how they leveraged auditory and visual information, all subjects exploited the reliability of unisensory cues. An alternative, nonprobabilistic model, in which subjects combined cues without regard for reliability, was not able to account for these trial-by-trial choices. These findings provide evidence that the brain relies on a probabilistic representation for numerosity decisions.

Young children bet on their numerical skills: metacognition in the numerical domain.

Metacognition, the ability to assess one's own knowledge, has been targeted as a critical learning mechanism in mathematics education. Yet the early childhood origins of metacognition have proven difficult to study. Using a novel nonverbal task and a comprehensive set of metacognitive measures, we provided the strongest evidence to date that young children are metacognitive. We showed that children as young as 5 years made metacognitive "bets" on their numerical discriminations in a wagering task. However, contrary to previous reports from adults, our results showed that children's metacognition is domain specific: Their metacognition in the numerical domain was unrelated to their metacognition in another domain (emotion discrimination). Moreover, children's metacognitive ability in only the numerical domain predicted their school-based mathematics knowledge. The data provide novel evidence that metacognition is a fundamental, domain-dependent cognitive ability in children. The findings have implications for theories of uncertainty and reveal new avenues for training metacognition in children.

Bayesian sampling in visual perception.

It is well-established that some aspects of perception and action can be understood as probabilistic inferences over underlying probability distributions. In some situations, it would be advantageous for the nervous system to sample interpretations from a probability distribution rather than commit to a particular interpretation. In this study, we asked whether visual percepts correspond to samples from the probability distribution over image interpretations, a form of sampling that we refer to as Bayesian sampling. To test this idea, we manipulated pairs of sensory cues in a bistable display consisting of two superimposed moving drifting gratings, and we asked subjects to report their perceived changes in depth ordering. We report that the fractions of dominance of each percept follow the multiplicative rule predicted by Bayesian sampling. Furthermore, we show that attractor neural networks can sample probability distributions if input currents add linearly and encode probability distributions with probabilistic population codes.

Natural language instructions induce compositional generalization in networks of neurons.

A fundamental human cognitive feat is to interpret linguistic instructions in order to perform novel tasks without explicit task experience. Yet, the neural computations that might be used to accomplish this remain poorly understood. We use advances in natural language processing to create a neural model of generalization based on linguistic instructions. Models are trained on a set of common psychophysical tasks, and receive instructions embedded by a pretrained language model. Our best models can perform a previously unseen task with an average performance of 83% correct based solely on linguistic instructions (that is, zero-shot learning). We found that language scaffolds sensorimotor representations such that activity for interrelated tasks shares a common geometry with the semantic representations of instructions, allowing language to cue the proper composition of practiced skills in unseen settings. We show how this model generates a linguistic description of a novel task it has identified using only motor feedback, which can subsequently guide a partner model to perform the task. Our models offer several experimentally testable predictions outlining how linguistic information must be represented to facilitate flexible and general cognition in the human brain.

The cost of accumulating evidence in perceptual decision making.

Decision making often involves the accumulation of information over time, but acquiring information typically comes at a cost. Little is known about the cost incurred by animals and humans for acquiring additional information from sensory variables due, for instance, to attentional efforts. Through a novel integration of diffusion models and dynamic programming, we were able to estimate the cost of making additional observations per unit of time from two monkeys and six humans in a reaction time (RT) random-dot motion discrimination task. Surprisingly, we find that the cost is neither zero nor constant over time, but for the animals and humans features a brief period in which it is constant but increases thereafter. In addition, we show that our theory accurately matches the observed reaction time distributions for each stimulus condition, the time-dependent choice accuracy both conditional on stimulus strength and independent of it, and choice accuracy and mean reaction times as a function of stimulus strength. The theory also correctly predicts that urgency signals in the brain should be independent of the difficulty, or stimulus strength, at each trial.

Predicting Bordeaux red wine origins and vintages from raw gas chromatograms.

Connecting chemical properties to various wine characteristics is of great interest to the science of olfaction as well as the wine industry. We explored whether Bordeaux wine chemical identities and vintages (harvest year) can be inferred from a common and affordable chemical analysis, namely, a combination of gas chromatography (GC) and electron ionization mass spectrometry. Using 12 vintages (within the 1990-2007 range) from 7 estates of the Bordeaux region, we report that, remarkably, nonlinear dimensionality reduction techniques applied to raw gas chromatograms recover the geography of the Bordeaux region. Using machine learning, we found that we can not only recover the estate perfectly from gas chromatograms, but also the vintage with up to 50% accuracy. Interestingly, we observed that the entire chromatogram is informative with respect to geographic location and age, thus suggesting that the chemical identity of a wine is not defined by just a few molecules but is distributed over a large chemical spectrum. This study demonstrates the remarkable potential of GC analysis to explore fundamental questions about the origin and age of wine.

Multi-timescale reinforcement learning in the brain.

To thrive in complex environments, animals and artificial agents must learn to act adaptively to maximize fitness and rewards. Such adaptive behavior can be learned through reinforcement learning1, a class of algorithms that has been successful at training artificial agents2-6 and at characterizing the firing of dopamine neurons in the midbrain7-9. In classical reinforcement learning, agents discount future rewards exponentially according to a single time scale, controlled by the discount factor. Here, we explore the presence of multiple timescales in biological reinforcement learning. We first show that reinforcement agents learning at a multitude of timescales possess distinct computational benefits. Next, we report that dopamine neurons in mice performing two behavioral tasks encode reward prediction error with a diversity of discount time constants. Our model explains the heterogeneity of temporal discounting in both cue-evoked transient responses and slower timescale fluctuations known as dopamine ramps. Crucially, the measured discount factor of individual neurons is correlated across the two tasks suggesting that it is a cell-specific property. Together, our results provide a new paradigm to understand functional heterogeneity in dopamine neurons, a mechanistic basis for the empirical observation that humans and animals use non-exponential discounts in many situations10-14, and open new avenues for the design of more efficient reinforcement learning algorithms.

Catalyzing next-generation Artificial Intelligence through NeuroAI.

Neuroscience has long been an essential driver of progress in artificial intelligence (AI). We propose that to accelerate progress in AI, we must invest in fundamental research in NeuroAI. A core component of this is the embodied Turing test, which challenges AI animal models to interact with the sensorimotor world at skill levels akin to their living counterparts. The embodied Turing test shifts the focus from those capabilities like game playing and language that are especially well-developed or uniquely human to those capabilities - inherited from over 500 million years of evolution - that are shared with all animals. Building models that can pass the embodied Turing test will provide a roadmap for the next generation of AI.

Marginalization in neural circuits with divisive normalization.

A wide range of computations performed by the nervous system involves a type of probabilistic inference known as marginalization. This computation comes up in seemingly unrelated tasks, including causal reasoning, odor recognition, motor control, visual tracking, coordinate transformations, visual search, decision making, and object recognition, to name just a few. The question we address here is: how could neural circuits implement such marginalizations? We show that when spike trains exhibit a particular type of statistics--associated with constant Fano factors and gain-invariant tuning curves, as is often reported in vivo--some of the more common marginalizations can be achieved with networks that implement a quadratic nonlinearity and divisive normalization, the latter being a type of nonlinear lateral inhibition that has been widely reported in neural circuits. Previous studies have implicated divisive normalization in contrast gain control and attentional modulation. Our results raise the possibility that it is involved in yet another, highly critical, computation: near optimal marginalization in a remarkably wide range of tasks.