GHL combines local Oja's rule with competitive learning and a global sign signal to outperform prior Hebbian methods and narrow the performance gap with backpropagation on large-scale tasks.
Arch VGG [33] ResNet [34] Layers 14 16 20 32 44 56 110 1202 Params 14.71 33.63 0.27 0.46 0.66 0.85 1.72 19.33 Acc 89.2989.48 86.72 86.86 87.09 87.17 87.21 86.97 Table 4
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.AI 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Hebbian Learning with Global Direction
GHL combines local Oja's rule with competitive learning and a global sign signal to outperform prior Hebbian methods and narrow the performance gap with backpropagation on large-scale tasks.