Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    INDEX
    Explanations

    removing redundancy

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     직접
    0.40
     ইহাই
    0.40
     melhores
    0.39
    pyx
    0.39
     अपरा
    0.39
     cep
    0.38
     نقص
    0.37
     dám
    0.37
     tông
    0.37
     सीधे
    0.36
    POSITIVE LOGITS
     unnecessary
    2.41
     superfluous
    2.13
     irrelevant
    2.05
     useless
    2.02
     pointless
    1.99
     inutile
    1.96
     needlessly
    1.94
     unnecessarily
    1.91
     अनावश्यक
    1.79
     meaningless
    1.73
    Act Density 0.074%

    No Known Activations