More Like this

1. Learning efficient task-dependent representations with synaptic plasticity

   https://doi.org/10.1101/2020.06.19.162172  

   Bredenberg, C.; Simoncelli, E.; Savin, C. (October 2020, Advances in Neural Information Processing Systems 33 (NeurIPS 2020))

   Neural populations encode the sensory world imperfectly: their capacity is limited by the number of neurons, availability of metabolic and other biophysical resources, and intrinsic noise. The brain is presumably shaped by these limitations, improving efficiency by discarding some aspects of incoming sensory streams, while preferentially preserving commonly occurring, behaviorally-relevant information. Here we construct a stochastic recurrent neural circuit model that can learn efficient, task-specific sensory codes using a novel form of reward-modulated Hebbian synaptic plasticity. We illustrate the flexibility of the model by training an initially unstructured neural network to solve two different tasks: stimulus estimation, and stimulus discrimination. The network achieves high performance in both tasks by appropriately allocating resources and using its recurrent circuitry to best compensate for different levels of noise. We also show how the interaction between stimulus priors and task structure dictates the emergent network representations. 

   more » « less
2. Modeling and dissociation of intrinsic and input-driven neural population dynamics underlying behavior

   https://doi.org/10.1073/pnas.2212887121  

   Vahidi, Parsa; Sani, Omid G; Shanechi, Maryam M (February 2024, Proceedings of the National Academy of Sciences)

   Neural dynamics can reflect intrinsic dynamics or dynamic inputs, such as sensory inputs or inputs from other brain regions. To avoid misinterpreting temporally structured inputs as intrinsic dynamics, dynamical models of neural activity should account for measured inputs. However, incorporating measured inputs remains elusive in joint dynamical modeling of neural-behavioral data, which is important for studying neural computations of behavior. We first show how training dynamical models of neural activity while considering behavior but not input or input but not behavior may lead to misinterpretations. We then develop an analytical learning method for linear dynamical models that simultaneously accounts for neural activity, behavior, and measured inputs. The method provides the capability to prioritize the learning of intrinsic behaviorally relevant neural dynamics and dissociate them from both other intrinsic dynamics and measured input dynamics. In data from a simulated brain with fixed intrinsic dynamics that performs different tasks, the method correctly finds the same intrinsic dynamics regardless of the task while other methods can be influenced by the task. In neural datasets from three subjects performing two different motor tasks with task instruction sensory inputs, the method reveals low-dimensional intrinsic neural dynamics that are missed by other methods and are more predictive of behavior and/or neural activity. The method also uniquely finds that the intrinsic behaviorally relevant neural dynamics are largely similar across the different subjects and tasks, whereas the overall neural dynamics are not. These input-driven dynamical models of neural-behavioral data can uncover intrinsic dynamics that may otherwise be missed. 

   more » « less
3. A Normative Theory for Causal Inference and Bayes Factor Computation in Neural Circuits

   Zhang, Wenhao; Wu, Si; Doiron, Brent; Lee, Tai Sing (January 2019, Advances in neural information processing systems)

   This study provides a normative theory for how Bayesian causal inference can be implemented in neural circuits. In both cognitive processes such as causal reasoning and perceptual inference such as cue integration, the nervous systems need to choose different models representing the underlying causal structures when making inferences on external stimuli. In multisensory processing, for example, the nervous system has to choose whether to integrate or segregate inputs from different sensory modalities to infer the sensory stimuli, based on whether the inputs are from the same or different sources. Making this choice is a model selection problem requiring the computation of Bayes factor, the ratio of likelihoods between the integration and the segregation models. In this paper, we consider the causal inference in multisensory processing and propose a novel generative model based on neural population code that takes into account both stimulus feature and stimulus reliability in the inference. In the case of circular variables such as heading direction, our normative theory yields an analytical solution for computing the Bayes factor, with a clear geometric interpretation, which can be implemented by simple additive mechanisms with neural population code. Numerical simulation shows that the tunings of the neurons computing Bayes factor are consistent with the "opposite neurons" discovered in dorsal medial superior temporal (MSTd) and the ventral intraparietal (VIP) areas for visual-vestibular processing. This study illuminates a potential neural mechanism for causal inference in the brain. 

   more » « less
4. Revealing nonlinear neural decoding by analyzing choices

   https://doi.org/10.1038/s41467-021-26793-9  

   Yang, Qianli; Walker, Edgar; Cotton, R. James; Tolias, Andreas S.; Pitkow, Xaq (December 2021, Nature Communications)

   Abstract Sensory data about most natural task-relevant variables are entangled with task-irrelevant nuisance variables. The neurons that encode these relevant signals typically constitute a nonlinear population code. Here we present a theoretical framework for quantifying how the brain uses or decodes its nonlinear information. Our theory obeys fundamental mathematical limitations on information content inherited from the sensory periphery, describing redundant codes when there are many more cortical neurons than primary sensory neurons. The theory predicts that if the brain uses its nonlinear population codes optimally, then more informative patterns should be more correlated with choices. More specifically, the theory predicts a simple, easily computed quantitative relationship between fluctuating neural activity and behavioral choices that reveals the decoding efficiency. This relationship holds for optimal feedforward networks of modest complexity, when experiments are performed under natural nuisance variation. We analyze recordings from primary visual cortex of monkeys discriminating the distribution from which oriented stimuli were drawn, and find these data are consistent with the hypothesis of near-optimal nonlinear decoding. 

   more » « less
5. Neural circuit regulation by identified modulatory projection neurons

   https://doi.org/10.3389/fnins.2023.1154769  

   Blitz, Dawn M. (March 2023, Frontiers in Neuroscience)

   Rhythmic behaviors (e.g., walking, breathing, and chewing) are produced by central pattern generator (CPG) circuits. These circuits are highly dynamic due to a multitude of input they receive from hormones, sensory neurons, and modulatory projection neurons. Such inputs not only turn CPG circuits on and off, but they adjust their synaptic and cellular properties to select behaviorally relevant outputs that last from seconds to hours. Similar to the contributions of fully identified connectomes to establishing general principles of circuit function and flexibility, identified modulatory neurons have enabled key insights into neural circuit modulation. For instance, while bath-applying neuromodulators continues to be an important approach to studying neural circuit modulation, this approach does not always mimic the neural circuit response to neuronal release of the same modulator. There is additional complexity in the actions of neuronally-released modulators due to: (1) the prevalence of co-transmitters, (2) local- and long-distance feedback regulating the timing of (co-)release, and (3) differential regulation of co-transmitter release. Identifying the physiological stimuli (e.g., identified sensory neurons) that activate modulatory projection neurons has demonstrated multiple “modulatory codes” for selecting particular circuit outputs. In some cases, population coding occurs, and in others circuit output is determined by the firing pattern and rate of the modulatory projection neurons. The ability to perform electrophysiological recordings and manipulations of small populations of identified neurons at multiple levels of rhythmic motor systems remains an important approach for determining the cellular and synaptic mechanisms underlying the rapid adaptability of rhythmic neural circuits. 

   more » « less