Recurrent network

Different stimuli elicit different transient neural responses in the brain. How is the information represented in the parallel neuronal activity of a neuronal population? We investigate this question in Neuropixel recordings of awake behaving mice by inferring the stimulus class (e.g., visual or tactile) from the activity. To quantify the separability, we use an optimally trained linear readout, which can be understood through the framework of Bayesian inference. We find that separability only depends on three geometric properties of the neuronal population vector, described by cones within th

…faster learning in feedforward-recurrent networks through a self-stabilizing feedback mechanism. We explore this hypothesis in the framework of learning novel stimulus representations in pre-structured recurren

…In this model a cerebral recurrent network receives feedback predictions from a cerebellar network, thereby decoupling learning in cerebral networks from future feedback. When trained in a simple sensorimotor task the model shows faster learning a

…interactions in which a cerebral recurrent network receives continuous cerebral feedback predictions from a cerebellar network, thereby decoupling learning in cerebral networks from future feedback. When trained in a simple sensorimotor task the model shows faster learning and reduced ataxia-like behaviours, in line with experimental observati

…tuned units also appear in artificial networks trained on cognitive tasks (Goh et al., 2021). It is often implicitly assumed that the emergence of such interpretable neurons plays a key role in the behavior of the trained network. Here we show that this is not necessarily the case. We train a biologically inspired artificial agent comprising two key components—a recurrent network and an associative memory—on a canonical rule

Predicting the future outcomes of actions forms the basis for reinforcement learning to shape goal-directed behaviours. Recent work showed that learning based on predicting future sensory experience using the current state and action of an agent leads to representations that resemble those in the brain, for instance place and grid cells of the medial temporal lobe. Here we ask if combining ongoing predictive learning of sensory events and of notional value of actions leading to rewards forms representations that enable more efficient goal-directed learning than reinforcement learning alone. Si

…learning in a feedforward PixelVAE network. Here we extend into the temporal domain, testing whether learning-by-prediction in a recurrent network discovers the properties of moving surfaces to which humans are sensitive. We rendered 10,000 close-up

…to argue that having an FeedForward Network (FN) responding like an recurrent network (RN) in a consciousness experiment is not enough to convince us the two are the same with regards to the posession of mental states and conscious experience. I will then argue that a robust functional equivalence between FFN and RN is akso not supported by the mathematical work on the Universal Approximator theorem, and is also unlikely to hold, as a conjecture, given data in cognitive neuroscience; I will argue that an equivalence of RN and FFN

…of non-spiking artificial neural networks. At the same time, spiking neural networks hold the promise for efficient processing of real-world sensory data by communicating using discrete events in continuous time. We derive the backpropagation algorithm for a recurrent network of spiking (leaky integrate-and-fire) neurons with hard thresholds and show that the backward dynamics amount to an event-based backpropagation of errors through time. Our derivation uses the jump condition

Homeostatic plasticity plays a key role in stabilizing neural network activity. But what is its role in neural information processing? We showed analytically how homeostasis changes collective dynamics and consequently information flow - depending on the input to the network. We then studied how input and homeostasis on a recurrent network of LIF neurons impacts information flow and task performance. We showed how we can tune the working point of the network, and found that, contrary to previous assumptions, there is not one optimal working point for a family of tasks, but each task may requir