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

    Neuronpedia

    INDEX
    Explanations

    problematic oroffending

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     ideales
    0.90
     Freude
    0.89
     Fähigkeiten
    0.89
    বন্ধু
    0.84
     ಆರೋಗ್ಯ
    0.83
     filosof
    0.83
     사회
    0.83
    myButtons
    0.81
     жизнь
    0.80
     социа
    0.80
    POSITIVE LOGITS
     offending
    2.02
     problematic
    1.78
     troublesome
    1.68
     culprit
    1.66
     unruly
    1.64
     faulty
    1.62
     culprits
    1.61
     offenders
    1.61
     violating
    1.61
     objectionable
    1.60
    Act Density 0.297%

    No Known Activations