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

    Neuronpedia

    INDEX
    Explanations

    legal and criminal offenses

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    ार
    1.10
    ל
    0.94
    ல்
    0.93
     これは
    0.89
    म
    0.89
     の
    0.87
     Serbia
    0.87
     crystall
    0.85
    ような
    0.84
    ॉ
    0.83
    POSITIVE LOGITS
    is
    1.23
    of
    0.99
    ,
    0.93
    iv
    0.91
    д
    0.90
    p
    0.88
    and
    0.86
    an
    0.84
    ue
    0.78
     offences
    0.77
    Act Density 0.004%

    No Known Activations