Response properties of sensory neurons are commonly described using receptive fields. This description may be formalized in a model that operates with small set linear filters whose outputs nonlinearly combined to determine the instantaneous firing rate. Spike-triggered average and covariance analyses can used estimate nonlinear combination rule from extracellular experimental data. We describe this methodology, demonstrating it simulated neuron examples emphasize practical issues arise situations.

Our ability to recognize objects despite large changes in position, size, and context is achieved through computations that are thought increase both the shape selectivity tolerance (“invariance”) of visual representation at successive stages ventral pathway [visual cortical areas V1, V2, V4 inferior temporal cortex (IT)]. However, these ideas have proven difficult test. Here, we consider how well population activity patterns two stream (V4 IT) discriminate between, generalize across,...

We propose a method that allows for rigorous statistical analysis of neural responses to natural stimuli are nongaussian and exhibit strong correlations. have in mind model which neurons selective small number stimulus dimensions out high-dimensional space, but within this subspace the can be arbitrarily nonlinear. Existing methods based on correlation functions between responses, these guaranteed work only case gaussian ensembles. As an alternative functions, we maximize mutual information...

Although popular accounts suggest that neurons along the ventral visual processing stream become increasingly selective for particular objects, this appears at odds with fact inferior temporal cortical (IT) are broadly tuned. To explore apparent contradiction, we compared in two stages (visual areas V4 and IT) rhesus macaque monkey. We confirmed IT indeed more conjunctions of features than increase feature conjunction selectivity is accompanied by an tolerance ("invariance") to...

Most accounts of image and object encoding in inferotemporal cortex (IT) focus on the distinct patterns spikes that different images evoke across IT population. By analyzing data collected from as monkeys performed a visual memory task, we demonstrate variation complementary coding scheme, magnitude population response, can largely account for how well will be remembered. To investigate origin memorability modulation, probed convolutional neural network models trained to categorize objects....

Our visual memory percepts of whether we have encountered specific objects or scenes before are hypothesized to manifest as decrements in neural responses inferotemporal cortex (IT) with stimulus repetition. To evaluate this proposal, recorded IT two monkeys performed a single-exposure task designed measure the rates forgetting time. We found that weighted linear read-out was better predictor monkeys’ and reaction time patterns than strict instantiation repetition suppression hypothesis,...

Several lines of evidence suggest that norepinephrine (NE) can modulate seizure activity. However, the experimental methods used in past cannot exclude possible role other neurotransmitters coreleased with NE from noradrenergic terminals. We have assessed susceptibility genetically engineered mice lack NE. Seizure was determined dopamine beta-hydroxylase null mutant (Dbh -/-) mouse using four different convulsant stimuli: 2,2,2-trifluroethyl ether (flurothyl), pentylenetetrazol (PTZ), kainic...

Although norepinephrine (NE) has been implicated in animal models of ethanol consumption for many years, the exact nature its influence is not clear. Lesioning and pharmacological studies examining role NE have yielded conflicting results. We took a genetic approach to determine effect depletion on ethanol-mediated behaviors by using dopamine β-hydroxylase knockout ( Dbh −/−) mice that specifically lack ability synthesize NE. −/− males reduced preference two-bottle choice paradigm show delay...

In addition to the role that our visual system plays in determining what we are seeing right now, computations contribute important ways predicting will see next. While of memory creating future predictions is often overlooked, efficient predictive computation requires use information about past estimate events. this article, introduce a framework for understanding relationship between and prediction review two classes mechanisms relies on create predictions. We also discuss principles...

Linear-nonlinear (LN) models and their extensions have proven successful in describing transformations from stimuli to spiking responses of neurons early stages sensory hierarchies. Neural at later are highly nonlinear generally been better characterized terms decoding performance on prespecified tasks. Here we develop a biologically plausible model for classification tasks, that refer as neural quadratic discriminant analysis (nQDA). Specifically, reformulate an optimal classifier LN-LN...