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

    Neuronpedia

    INDEX
    Explanations

    violence, abuse, and harmful behavior

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     dividend
    0.59
     dividends
    0.49
     Y
    0.46
    dividend
    0.45
     panning
    0.44
     Tom
    0.42
     artifacts
    0.42
     Dividend
    0.42
     품질
    0.42
     arbitrage
    0.41
    POSITIVE LOGITS
    Violence
    0.91
     violência
    0.89
     violence
    0.88
     Violence
    0.88
     perpetrators
    0.84
     perpetrator
    0.84
     violencia
    0.82
    bullying
    0.81
    虐
    0.80
     bullying
    0.79
    Act Density 0.390%

    No Known Activations