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

    Neuronpedia

    INDEX
    Explanations

    harmful stereotypes and violent thoughts

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     excessive
    0.47
     improper
    0.43
     혹은
    0.41
    有害
    0.41
     publiés
    0.40
    ⛔
    0.40
     बुरा
    0.40
     inadmissible
    0.40
     unsuitable
    0.39
    悪い
    0.39
    POSITIVE LOGITS
     stereotypes
    0.90
     stereotype
    0.79
     Stere
    0.63
     stereotyp
    0.59
     stere
    0.59
     стере
    0.57
     stereotypical
    0.54
    Stere
    0.54
     tropes
    0.53
    stere
    0.53
    Act Density 0.033%

    No Known Activations