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

    Neuronpedia

    INDEX
    Explanations

    speech attribution verbs

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    使用的
    0.31
    Ov
    0.29
    潭
    0.28
    ）。
    0.28
    тину
    0.28
    处于
    0.27
    ERN
    0.27
    使其
    0.27
     الحصول
    0.27
    ٹرول
    0.26
    POSITIVE LOGITS
     says
    0.74
    says
    0.67
     exclaimed
    0.62
     said
    0.58
     säger
    0.57
     zegt
    0.56
     groaned
    0.55
     cried
    0.54
     saith
    0.54
     Says
    0.53
    Act Density 0.003%

    No Known Activations