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

    Neuronpedia

    INDEX
    Explanations

    specific language that expresses condemnation or hostility towards groups of people, particularly reflecting themes of racism and discrimination

    oai_token-act-pair · gpt-4o-miniTriggered by @bot

    derogatory insults

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     الرياضيه
    -0.53
     genial
    -0.47
    iastes
    -0.46
    +#+#
    -0.44
    OGND
    -0.44
    UIFont
    -0.42
    sidemargin
    -0.42
    ArrowToggle
    -0.42
     involunt
    -0.41
    AsUp
    -0.41
    POSITIVE LOGITS
     stupid
    0.67
     smelly
    0.60
     pesky
    0.59
    stupid
    0.58
     foreigners
    0.56
     foreigner
    0.56
     inferior
    0.55
     barbaric
    0.55
     useless
    0.54
     extranjera
    0.54
    Act Density 0.118%

    No Known Activations