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

    Neuronpedia

    INDEX
    Explanations

    shall, may, liable

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     sultry
    0.52
     murderous
    0.48
     sweaty
    0.46
    ຫຼາຍ
    0.45
     전략
    0.43
     پھی
    0.42
     hedon
    0.41
    が出て
    0.41
     grungy
    0.41
     vraiment
    0.41
    POSITIVE LOGITS
     shall
    0.79
    reasonable
    0.50
     practicable
    0.47
     Whenever
    0.47
     Shall
    0.46
    shall
    0.46
     upon
    0.45
     hereby
    0.45
     appropriate
    0.45
    provide
    0.45
    Act Density 0.043%

    No Known Activations