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

    Neuronpedia

    INDEX
    Explanations

    words beginning with ish or mish

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     withstand
    3.34
    その
    3.23
    そして
    3.12
     bole
    3.08
    على
    2.96
    いずれ
    2.95
    もちろん
    2.94
    tır
    2.84
     ihop
    2.79
    특별시
    2.78
    POSITIVE LOGITS
    на
    3.84
    as
    3.48
    ம்
    3.15
    in
    3.09
    ы
    3.07
    लिये
    2.98
    ਤੇ
    2.96
    ness
    2.95
    ා
    2.90
    at
    2.79
    Act Density 0.015%

    No Known Activations