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

    Neuronpedia

    INDEX
    Explanations

    exploits, notation, dependent, passion

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    are
    0.91
    0
    0.86
    কে
    0.77
    ang
    0.76
    ка
    0.75
    ↵
    0.72
    <0x80>
    0.72
    na
    0.72
    nante
    0.68
    ok
    0.67
    POSITIVE LOGITS
    ى
    0.82
     powied
    0.80
    ۸
    0.77
    шымта
    0.77
    ૭
    0.76
     significativa
    0.75
    显得
    0.73
     PROVID
    0.73
    远的
    0.73
    ༦
    0.73
    Act Density 0.001%

    No Known Activations