All articles tagged with #brain computation
Researchers using voltage-imaging in mice show dendrites can compute independently of the cell body, effectively turning a neuron into a multi-processor. In novel environments, some dendrites encode new locations before the soma and can preserve past activity as memory traces, suggesting dendrites act as local memory devices and that brain power is much greater than previously thought due to complex dendritic architectures.
The article proposes biological computationalism, a new framework suggesting that consciousness arises from the brain's unique hybrid, scale-inseparable, and energy-constrained physical computation, challenging traditional views of cognition and highlighting limitations of current AI models.
A study led by Dr. Richard Naud at the University of Ottawa reveals how dendrites, the neuron's antennas, control the variability in neuronal responses, offering critical insights into biological computation with implications for both neuroscience and AI development. The research provides a mathematical model for simulating neuronal networks with active dendrites, shedding light on the role of dendritic input in shaping the response of neuronal ensembles and offering potential applications in understanding learning, working memory, and theories of attractor states.
A new study challenges the long-standing hypothesis in neuroscience that the soma (cell body) of neurons is the primary computational element in the brain, suggesting that dendrites also play a crucial role. Researchers conducted experiments under non-physiological conditions and found that dendrites control neuron features such as firing frequency and stimulation threshold. This discovery could reshape our understanding of brain states, learning processes, and degenerative diseases. The results call for a re-examination of the origin of degenerative diseases and question the traditional attribution of awake and sleep states to the soma.