52% Yes, a signiicant crisis 3% No, there is no crisis 7% Don't know 38% Yes, a slight crisis 38% Yes, a slight crisis 1,576 RESEARCHERS SURVEYED M ore than 70% of researchers have tried and failed to reproduce another scientist's experiments, and more than half have failed to reproduce their own experiments. Those are some of the telling figures that emerged from Nature's survey of 1,576 researchers who took a brief online questionnaire on reproducibility in research. The data reveal sometimes-contradictory attitudes towards reproduc-ibility. Although 52% of those surveyed agree that there is a significant 'crisis' of reproducibility, less than 31% think that failure to reproduce published results means that the result is probably wrong, and most say that they still trust the published literature. Data on how much of the scientific literature is reproducible are rare and generally bleak. The best-known analyses, from psychology 1 and cancer biology 2 , found rates of around 40% and 10%, respectively. Our survey respondents were more optimistic: 73% said that they think that at least half of the papers in their field can be trusted, with physicists and chemists generally showing the most confidence. The results capture a confusing snapshot of attitudes around these issues, says Arturo Casadevall, a microbiologist at the Johns Hopkins Bloomberg School of Public Health in Baltimore, Maryland. "At the current time there is no consensus on what reproducibility is or should be. " But just recognizing that is a step forward, he says. "The next step may be identifying what is the problem and to get a consensus. "

Neuronal activity in the brain is variable, yet both perception and behavior are generally reliable. How does the brain achieve this? Here, we show that the conjunctive coding of multiple stimulus features, commonly known as nonlinear mixed selectivity, may be used by the brain to support reliable information transmission using unreliable neurons. Nonlinearly mixed feature representations have been observed throughout primary sensory, decisionmaking, and motor brain areas. In these areas, different features are almost always nonlinearly mixed to some degree, rather than represented separately or with only additive (linear) mixing, which we refer to as pure selectivity. Mixed selectivity has been previously shown to support flexible linear decoding for complex behavioral tasks. Here, we show that it has another important benefit: in many cases, it makes orders of magnitude fewer decoding errors than pure selectivity even when both forms of selectivity use the same number of spikes. This benefit holds for sensory, motor, and more abstract, cognitive representations. Further, we show experimental evidence that mixed selectivity exists in the brain even when it does not enable behaviorally useful linear decoding. This suggests that nonlinear mixed selectivity may be a general coding scheme exploited by the brain for reliable and efficient neural computation. Author summaryNeurons in the brain are unreliable, while both perception and behavior are generally reliable. In this work, we study how the neural population response to sensory, motor, and cognitive features can produce this reliability. Across the brain, single neurons have been shown to respond to particular conjunctions of multiple features, termed nonlinear mixed selectivity. In this work, we show that populations of these mixed selective neurons lead to many fewer decoding errors than populations without mixed selectivity, even when both neural codes are given the same number of spikes. We show that the reliability benefits from mixed selectivity are quite general, holding under different assumptions about PLOS Computational Biology | https://doi.

Neuronal activity in the brain is variable, yet both perception and behavior are generally reliable. How does the brain achieve this? Here, we show that the conjunctive coding of multiple stimulus features, commonly known as nonlinear mixed selectivity, may be used by the brain to support reliable information transmission using unreliable neurons. Nonlinear mixed selectivity (NMS) has been observed widely across the brain, from primary sensory to decision-making to motor areas. Representations of stimulus features are nearly always mixed together, rather than represented separately or with only additive (linear) mixing, as in pure selectivity. NMS has been previously shown to support flexible linear decoding for complex behavioral tasks. Here, we show that NMS has another important benefit: it requires as little as half the metabolic energy required by pure selectivity to achieve the same level of transmission reliability. This benefit holds for sensory, motor, and more abstract, cognitive representations. Further, we show experimental evidence that NMS exists in the brain even when it does not enable behaviorally useful linear decoding. This suggests that NMS may be a general coding scheme exploited by the brain for reliable and efficient neural computation.

Humans and other animals demonstrate a remarkable ability to generalize knowledge across distinct contexts and objects during natural behavior. We posit that this ability to generalize arises from a specific representational geometry, that we call abstract and that is referred to as disentangled in machine learning. These abstract representations have been observed in recent neurophysiological studies. However, it is unknown how they emerge. Here, using feedforward neural networks, we demonstrate that the learning of multiple tasks causes abstract representations to emerge, using both supervised and reinforcement learning. We show that these abstract representations enable few-sample learning and reliable generalization on novel tasks. We conclude that abstract representations of sensory and cognitive variables may emerge from the multiple behaviors that animals exhibit in the natural world, and, as a consequence, could be pervasive in high-level brain regions. We also make several specific predictions about which variables will be represented abstractly.

Humans and other animals demonstrate a remarkable ability to generalize knowledge across distinct contexts and objects during natural behavior. We posit that this ability depends on the geometry of the neural population representations of these objects and contexts. Specifically, abstract, or disentangled, neural representations -- in which neural population activity is a linear function of the variables important for making a decision -- are known to allow for this kind of generalization. Further, recent neurophysiological studies have shown that the brain has sufficiently abstract representations of some sensory and cognitive variables to enable generalization across distinct contexts. However, it is unknown how these abstract representations emerge. Here, using feedforward neural networks, we demonstrate a simple mechanism by which these abstract representations can be produced: The learning of multiple distinct classification tasks. We demonstrate that, despite heterogeneity in the task structure, abstract representations that enable reliable generalization can be produced from a variety of different inputs -- including standard nonlinearly mixed inputs, inputs that mimic putative representations from early sensory areas, and even simple image inputs from a standard machine learning data set. Thus, we conclude that abstract representations of sensory and cognitive variables emerge from the multiple behaviors that animals exhibit in the natural world, and may be pervasive in high-level brain regions. We make several specific predictions about which variables will be represented abstractly as well as show how these representations can be detected.