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

    Neuronpedia

    INDEX
    Explanations

    all can

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     that
    0.94
     and
    0.91
     is
    0.91
     que
    0.90
     और
    0.86
     was
    0.86
     arch
    0.85
     person
    0.84
     witch
    0.84
     potato
    0.83
    POSITIVE LOGITS
    u
    1.30
     เงี้ย
    1.15
     گے۔
    1.10
    Đây
    1.09
    é
    1.02
    éléments
    1.02
    ا۔
    0.99
    。
    0.97
    iciones
    0.96
     এছাড়াও
    0.96
    Act Density 0.457%

    No Known Activations