Attention - Modulating cortical state locally
How do we handle the vast stream of inputs—sights, sounds, smells—continuously impinging on our senses?
How do we handle the vast stream of inputs—sights, sounds, smells—continuously impinging on our senses?
Neurogrid is part of a profound shift in computing, away from the sequential, step-by-step Von Neumann architecture towards a parallel, interconnected architecture more like the brain.
To map computations onto Neurogrid, we use the Neural Engineering Framework.
Computers' inherent limitations as neural simulation platforms are addressed by neuromorphic chips: Their fundamental component is not a logic gate but a silicon neuron.
Barn owls catch mice scurrying in the dark by telling which ear sound arrives at first—to within a microsecond!
The genome's billion (10^9) bits cannot specify where each synapse in the brain goes—--a hundred million times more bits are required to list the quadrillion (10^15) connections among the brain's trillion (10^12) neurons
Creating and simulating cortical models on Neurogrid is straightforward--you describe your neural model by writing a Python script.
Spaun autonomously performs several different cognitive visuomotor tasks but runs 9,000 times slower than real-time on a 16-core PC.
How does information persist stably in an inherently noisy brain without deterioration, over timescales of seconds to minutes?
In their quest to understand how cognition emerges at the behavioral level from mechanisms at the molecular level, researchers have pursued three approaches: Experiment, Theory, and Computation.
The retina’s ganglion cells, whose signals the optic nerve carries, only interpret a small portion of the visual scene for the brain.
Recent breakthroughs in neuromorphic engineering make it possible to combine analog's parallel operation with digital's programmability, reaping the best of both worlds.
BrainGate's partially implanted brain machine interface (BMI) uses half of the 100 mW it consumes to wirelessly transmit neural signals recorded on 100 microelectrodes.
A computer that executes one instruction every time the brain activates a synapse would consume a hundred megawatts!
When we visit a new place, our brains remember it. The brain region responsible is the hippocampus, which stores sensory experiences.
How can a neuron hear anything over the cacophony created by ten thousand inputs?