INDEX
    Explanations

    programming, saving old values

    the neuron activates on code or text that saves or backs up an original value or state (e.g. “save original…,” “old…,” then later restore) before making changes.

    New Auto-Interp
    Negative Logits
     oppos
    -0.08
    &view
    -0.07
     Manifest
    -0.07
    ernity
    -0.07
     ambos
    -0.07
    ifié
    -0.07
     Operand
    -0.06
    -being
    -0.06
     wei
    -0.06
     Germany
    -0.06
    POSITIVE LOGITS
    olate
    0.07
    IDA
    0.06
     =========================================================================
    0.06
    TRY
    0.06
    093
    0.06
    -dd
    0.06
     regulate
    0.06
    _TD
    0.05
    655
    0.05
    ышлен
    0.05
    Act Density 0.063%

    No Known Activations