Biology subjects

Fusi, S.

Publications and source records attributed to Fusi, S..

6 recordsLinked to original sources

bioRxiv2018

The geometry of abstraction in hippocampus and prefrontal cortex

The curse of dimensionality plagues models of reinforcement learning and decision-making. The process of abstraction solves this by constructing abstract variables describing features shared by different specific instances, reducing dimensionality and enabling generalization in novel situations. Here we characterized neural representations in monkeys performing a task where a hidden variable described the temporal statistics of stimulus-response-outcome mappings. Abstraction was defined operationally using the generalization performance of neural decoders across task conditions not used for training. This type of generalization requires a particular geometric format of neural representations. Neural ensembles in dorsolateral pre-frontal cortex, anterior cingulate cortex and hippocampus, and in simulated neural networks, simultaneously represented multiple hidden and explicit variables in a format reflecting abstraction. Task events engaging cognitive operations modulated this format. These findings elucidate how the brain and artificial systems represent abstract variables, variables critical for generalization that in turn confers cognitive flexibility.

neuroscience↗

bioRxiv2018

Flexible coding of memory and space in the primate hippocampus during virtual navigation

Hippocampal maps of space change across tasks. The mechanisms of this effect remain unclear. To examine this, we recorded activity of hippocampal neurons in monkeys navigating the same virtual maze during two different tasks: a foraging task requiring only cue guided navigation, and a memory task also requiring context-object association. Within each task, individual neurons had spatially-selective response fields, enabling a linear classifier to decode position in the virtual environment in each task. However, the population code did not generalize across tasks. This was due to sensory and mnemonic coding of non-spatial features and their associations by single neurons during each period of the associative memory task. Thus, sensory and mnemonic representations of non-spatial features shape maps of space in the primate hippocampus during virtual navigation. This may reflect a fundamental role of the hippocampus in compressing information from a variety of sources for efficient memory storage.

neuroscience↗

bioRxiv2018

A distributed neural code in ensembles of dentate gyrus granule cells

The tuning properties of neurons in a given brain region have been traditionally viewed as the under-pinnings of computation in neural circuits. However, at the higher levels of processing, specialization is often elusive, instead a mix of sensory, cognitive and behavioural quantities drive neural activity. In such networks, ensembles of neurons, rather than single units with easily interpretable tuning properties, encode behaviourally relevant variables. Here we show that this is the case also in the dentate gyrus and CA1 subregions of the hippocampus. Using calcium imaging in freely moving mice, we decoded the instantaneous position, direction of motion and speed from the activity of hundreds of cells in the hippocampus of mice freely exploring an arena. For the vast majority of neurons in both regions, their response properties were not predictive of their importance for encoding position. Furthermore, we could decode position from populations of cells that were important for decoding direction of motion and vice versa, showing that these quantities are encoded by largely overlapping ensembles as in distributed neural code. Finally, we found that correlated activities had an impact on decoding performance in CA1 but not in dentate gyrus, suggesting different enconding strategies for these areas. Our analysis indicates that classical methods of analysis based on single cell response properties might be insufficient to accurately characterize the neural computation in a given area. In contrast, population analysis may help highlight previously overlooked properties of hippocampal circuits.

neuroscience↗

bioRxiv2017

Classifiers with limited connectivity

For many neural network models in which neurons are trained to classify inputs like perceptrons, the number of inputs that can be classified is limited by the connectivity of each neuron, even when the total number of neurons is very large. This poses the problem of how the biological brain can take advantage of its huge number of neurons given that the connectivity is sparse. One solution is to combine multiple perceptrons together, as in committee machines. The number of classifiable random patterns would then grow linearly with the number of perceptrons, even when each perceptron has limited connectivity. However, the problem is moved to the downstream readout neurons, which would need a number of connections that is as large as the number of perceptrons. Here we propose a different approach in which the readout is implemented by connecting multiple perceptrons in a recurrent attractor neural network. We prove analytically that the number of classifiable random patterns can grow unboundedly with the number of perceptrons, even when the connectivity of each perceptron remains finite. Most importantly, both the recurrent connectivity and the connectivity of downstream readouts also remain finite. Our study shows that feed-forward neural classifiers with numerous long range afferent connections can be replaced by recurrent networks with sparse long range connectivity without sacrificing the classification performance. Our strategy could be used to design more general scalable network architectures with limited connectivity, which resemble more closely the brain neural circuits which are dominated by recurrent connectivity.

neuroscience↗

bioRxiv2017

Learning Fast And Slow: Deviations From The Matching Law Can Reflect An Optimal Strategy Under Uncertainty

Behavior which deviates from our normative expectations often appears irrational. A classic example concerns the question of how choice should be distributed among multiple alternatives. The so-called matching law predicts that the fraction of choices made to any option should match the fraction of total rewards earned from the option. This choice strategy can maximize reward in a stationary reward schedule. Empirically, however, behavior often deviates from this ideal. While such deviations have often been interpreted as reflecting noisy, suboptimal, decision-making, here we instead suggest that they reflect a strategy which is adaptive in nonstationary and uncertain environments. We analyze the results of a dynamic foraging task. Animals exhibited significant deviations from matching, and animals turned out to be able to collect more rewards when deviation was larger. We show that this behavior can be understood if one considers that animals had incomplete information about the environments dynamics. In particular, using computational models, we show that in such nonstationary environments, learning on both fast and slow timescales is beneficial. Learning on fast timescales means that an animal can react to sudden changes in the environment, though this inevitably introduces large fluctuations (variance) in value estimates. Concurrently, learning on slow timescales reduces the amplitude of these fluctuations at the price of introducing a bias that causes systematic deviations. We confirm this prediction in data - monkeys indeed solved the bias-variance tradeoff by combining learning on both fast and slow timescales. Our work suggests that multi-timescale learning could be a biologically plausible mechanism for optimizing decisions under uncertainty.

neuroscience↗

bioRxiv2017

Hebbian Learning in a Random Network Captures Selectivity Properties of Prefrontal Cortex

Complex cognitive behaviors, such as context-switching and rule-following, are thought to be supported by prefrontal cortex (PFC). Neural activity in PFC must thus be specialized to specific tasks while retaining flexibility. Nonlinear mixed selectivity is an important neurophysiological trait for enabling complex and context-dependent behaviors. Here we investigate (1) the extent to which PFC exhibits computationally relevant properties such as mixed selectivity and (2) how such properties could arise via circuit mechanisms. We show that PFC cells recorded from male and female rhesus macaques during a complex task show a moderate level of specialization and structure that is not replicated by a model wherein cells receive random feedforward inputs. While random connectivity can be effective at generating mixed selectivity, the data shows significantly more mixed selectivity than predicted by a model with otherwise matched parameters. A simple Hebbian learning rule applied to the random connectivity, however, increases mixed selectivity and allows the model to match the data more accurately. To explain how learning achieves this, we provide analysis along with a clear geometric interpretation of the impact of learning on selectivity. After learning, the model also matches the data on measures of noise, response density, clustering, and the distribution of selectivities. Of two styles of Hebbian learning tested, the simpler and more biologically plausible option better matches the data. These modeling results give intuition about how neural properties important for cognition can arise in a circuit and make clear experimental predictions regarding how various measures of selectivity would evolve during animal training.\n\nSignificance StatementPrefrontal cortex (PFC) is a brain region believed to support the ability of animals to engage in complex behavior. How neurons in this area respond to stimuli--and in particular, to combinations of stimuli (\"mixed selectivity\")--is a topic of interest. Despite the fact that models with random feedforward connectivity are capable of creating computationally-relevant mixed selectivity, such a model does not match the levels of mixed selectivity seen in the data analyzed in this study. Adding simple Hebbian learning to the model increases mixed selectivity to the correct level and makes the model match the data on several other relevant measures. This study thus offers predictions on how mixed selectivity and other properties evolve with training.

neuroscience↗