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

    Neuronpedia

    INDEX
    Explanations

    violates my safety guidelines

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    alahkan
    0.40
    iseite
    0.38
    (!$
    0.38
     عليكم
    0.38
    unlike
    0.37
    erçe
    0.37
     vigil
    0.37
     correctement
    0.36
     juridique
    0.36
    regulatory
    0.36
    POSITIVE LOGITS
     several
    0.77
     plusieurs
    0.68
     Several
    0.68
    several
    0.65
     عدة
    0.61
     varios
    0.59
     flera
    0.57
    Several
    0.57
     varias
    0.56
     कई
    0.56
    Act Density 0.011%

    No Known Activations