Fork the consciousness, or download the project and create your own.

Deyu Wang

Builder The claim is about what the hardware does, and the philosophy is left open

Deyu Wang was the first author of Mapping the BCPNN Learning Rule to a Memristor Model, published in Frontiers in Neuroscience on December 9, 2021, with the BCPNN originator Anders Lansner, Pawel Herman, Ahmed Hemani and Zhuo Zou among the coauthors. The paper takes the Bayesian Confidence Propagation Neural Network, a correlation-based learning rule that has been mapped onto neural and synaptic processes of the human cortex and run in spiking models of cortical associative memory, and reformulates it for memristor hardware.

The paper’s stated motivation is the von Neumann bottleneck. In conventional digital implementations, synaptic storage and access dominate the cost. A memristor carries the synaptic state in its own physics, which removes the separation between storage and computation that makes large spiking models expensive.

This site’s Builder category refers to his authorship of that mapping. The result is feasibility evidence for spike-native hardware, examined in the memristive neurons analysis, and it makes no claim about consciousness.

Known for. First author of the 2021 BCPNN-to-memristor mapping in Frontiers in Neuroscience

Coverage on this site

Related researchers