A proposed ReRAM synapse stores the eligibility trace of the e-prop learning rule as local temperature, with weight updates driven by temperature-dependent conductance change, but the stated thermal time constant is too short to accumulate the trace across the milli-second training frames.
Title resolution pending
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.ET 1years
2024 1verdicts
REJECT 1representative citing papers
citing papers explorer
-
NeoHebbian Synapses to Accelerate Online Training of Neuromorphic Hardware
A proposed ReRAM synapse stores the eligibility trace of the e-prop learning rule as local temperature, with weight updates driven by temperature-dependent conductance change, but the stated thermal time constant is too short to accumulate the trace across the milli-second training frames.