Specifically, for BEAN-1, We can see a clear trend where the model performance drops when α is too small or too big.
← all excerpts
BEAN: Interpretable and Efficient Learning With Biologically-Enhanced Artificial Neuronal Assembly Regularization.
1
—
—