Grounded language acquisition through the eyes and ears of a single child.

Starting around 6 to 9 months of age, children begin acquiring their first words, linking spoken words to their visual counterparts. How much of this knowledge is learnable from sensory input with relatively generic learning mechanisms, and how much requires stronger inductive biases? Using longitudinal head-mounted camera recordings from one child aged 6 to 25 months, we trained a relatively generic neural network on 61 hours of correlated visual-linguistic data streams, learning feature-based representations and cross-modal associations. Our model acquires many word-referent mappings present in the child's everyday experience, enables zero-shot generalization to new visual referents, and aligns its visual and linguistic conceptual systems. These results show how critical aspects of grounded word meaning are learnable through joint representation and associative learning from one child's input.

Spatial relation categorization in infants and deep neural networks.

Spatial relations, such as above, below, between, and containment, are important mediators in children's understanding of the world (Piaget, 1954). The development of these relational categories in infancy has been extensively studied (Quinn, 2003) yet little is known about their computational underpinnings. Using developmental tests, we examine the extent to which deep neural networks, pretrained on a standard vision benchmark or egocentric video captured from one baby's perspective, form categorical representations for visual stimuli depicting relations. Notably, the networks did not receive any explicit training on relations. We then analyze whether these networks recover similar patterns to ones identified in development, such as reproducing the relative difficulty of categorizing different spatial relations and different stimulus abstractions. We find that the networks we evaluate tend to recover many of the patterns observed with the simpler relations of "above versus below" or "between versus outside", but struggle to match developmental findings related to "containment". We identify factors in the choice of model architecture, pretraining data, and experimental design that contribute to the extent the networks match developmental patterns, and highlight experimental predictions made by our modeling results. Our results open the door to modeling infants' earliest categorization abilities with modern machine learning tools and demonstrate the utility and productivity of this approach.

Publisher Correction: A diverse range of factors affect the nature of neural representations underlying short-term memory.

The original and corrected figures are shown in the accompanying Publisher Correction.

A diverse range of factors affect the nature of neural representations underlying short-term memory.

Sequential and persistent activity models are two prominent models of short-term memory in neural circuits. In persistent activity models, memories are represented in persistent or nearly persistent activity patterns across a population of neurons, whereas in sequential models, memories are represented dynamically by a sequential activity pattern across the population. Experimental evidence for both models has been reported previously. However, it has been unclear under what conditions these two qualitatively different types of solutions emerge in neural circuits. Here, we address this question by training recurrent neural networks on several short-term memory tasks under a wide range of circuit and task manipulations. We show that both sequential and nearly persistent solutions are part of a spectrum that emerges naturally in trained networks under different conditions. Our results help to clarify some seemingly contradictory experimental results on the existence of sequential versus persistent activity-based short-term memory mechanisms in the brain.

Efficient probabilistic inference in generic neural networks trained with non-probabilistic feedback.

Animals perform near-optimal probabilistic inference in a wide range of psychophysical tasks. Probabilistic inference requires trial-to-trial representation of the uncertainties associated with task variables and subsequent use of this representation. Previous work has implemented such computations using neural networks with hand-crafted and task-dependent operations. We show that generic neural networks trained with a simple error-based learning rule perform near-optimal probabilistic inference in nine common psychophysical tasks. In a probabilistic categorization task, error-based learning in a generic network simultaneously explains a monkey's learning curve and the evolution of qualitative aspects of its choice behavior. In all tasks, the number of neurons required for a given level of performance grows sublinearly with the input population size, a substantial improvement on previous implementations of probabilistic inference. The trained networks develop a novel sparsity-based probabilistic population code. Our results suggest that probabilistic inference emerges naturally in generic neural networks trained with error-based learning rules.Behavioural tasks often require probability distributions to be inferred about task specific variables. Here, the authors demonstrate that generic neural networks can be trained using a simple error-based learning rule to perform such probabilistic computations efficiently without any need for task specific operations.

Neural population coding of multiple stimuli.

In natural scenes, objects generally appear together with other objects. Yet, theoretical studies of neural population coding typically focus on the encoding of single objects in isolation. Experimental studies suggest that neural responses to multiple objects are well described by linear or nonlinear combinations of the responses to constituent objects, a phenomenon we call stimulus mixing. Here, we present a theoretical analysis of the consequences of common forms of stimulus mixing observed in cortical responses. We show that some of these mixing rules can severely compromise the brain's ability to decode the individual objects. This cost is usually greater than the cost incurred by even large reductions in the gain or large increases in neural variability, explaining why the benefits of attention can be understood primarily in terms of a stimulus selection, or demixing, mechanism rather than purely as a gain increase or noise reduction mechanism. The cost of stimulus mixing becomes even higher when the number of encoded objects increases, suggesting a novel mechanism that might contribute to set size effects observed in myriad psychophysical tasks. We further show that a specific form of neural correlation and heterogeneity in stimulus mixing among the neurons can partially alleviate the harmful effects of stimulus mixing. Finally, we derive simple conditions that must be satisfied for unharmful mixing of stimuli.

Toward ecologically realistic theories in visual short-term memory research.

Recent evidence from neuroimaging and psychophysics suggests common neural and representational substrates for visual perception and visual short-term memory (VSTM).Visual perception is adapted to a rich set of statistical regularities present in the natural visual environment. Common neural and representational substrates for visual perception and VSTM suggest that VSTM is adapted to these same statistical regularities too. This article discusses how the study of VSTM can be extended to stimuli that are ecologically more realistic than those commonly used in standard VSTM experiments and what the implications of such an extension could be for our current view of VSTM. We advocate for the development of unified models of visual perception and VSTM­probabilistic and hierarchical in nature­ incorporating prior knowledge of natural scene statistics.

A probabilistic clustering theory of the organization of visual short-term memory.

Experimental evidence suggests that the content of a memory for even a simple display encoded in visual short-term memory (VSTM) can be very complex. VSTM uses organizational processes that make the representation of an item dependent on the feature values of all displayed items as well as on these items' representations. Here, we develop a probabilistic clustering theory (PCT) for modeling the organization of VSTM for simple displays. PCT states that VSTM represents a set of items in terms of a probability distribution over all possible clusterings or partitions of those items. Because PCT considers multiple possible partitions, it can represent an item at multiple granularities or scales simultaneously. Moreover, using standard probabilistic inference, it automatically determines the appropriate partitions for the particular set of items at hand and the probabilities or weights that should be allocated to each partition. A consequence of these properties is that PCT accounts for experimental data that have previously motivated hierarchical models of VSTM, thereby providing an appealing alternative to hierarchical models with prespecified, fixed structures. We explore both an exact implementation of PCT based on Dirichlet process mixture models and approximate implementations based on Bayesian finite mixture models. We show that a previously proposed 2-level hierarchical model can be seen as a special case of PCT with a single cluster. We show how a wide range of previously reported results on the organization of VSTM can be understood in terms of PCT. In particular, we find that, consistent with empirical evidence, PCT predicts biases in estimates of the feature values of individual items and also predicts a novel form of dependence between estimates of the feature values of different items. We qualitatively confirm this last prediction in 3 novel experiments designed to directly measure biases and dependencies in subjects' estimates.

Visual learning with reliable and unreliable features.

Existing studies of sensory integration demonstrate how the reliabilities of perceptual cues or features influence perceptual decisions. However, these studies tell us little about the influence of feature reliability on visual learning. In this article, we study the implications of feature reliability for perceptual learning in the context of binary classification tasks. We find that finite sets of training data (i.e., the stimuli and corresponding class labels used on training trials) contain different information about a learner's parameters associated with reliable versus unreliable features. In particular, the statistical information provided by a finite number of training trials strongly constrains the set of possible parameter values associated with unreliable features, but only weakly constrains the parameter values associated with reliable features. Analyses of human subjects' performances reveal that subjects were sensitive to this statistical information. Additional analyses examine why subjects were sub-optimal visual learners.