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

    Neuronpedia

    INDEX
    Explanations

    mentions of racism, harmful/discriminatory content, or policy-style refusals explaining why hateful content can't be provided.

    oai_token-act-pair · gpt-5-miniTriggered by @drejcpesjakpesjak
    New Auto-Interp
    Negative Logits
     Unexpected
    0.87
     Graphics
    0.83
     الماء
    0.79
    Unexpected
    0.78
    InnoDB
    0.78
    ከናወ
    0.78
    Graphics
    0.77
    Physics
    0.76
     Acrobat
    0.76
     शॉट
    0.76
    POSITIVE LOGITS
     perpetuated
    1.83
     patriarchal
    1.76
     dehuman
    1.75
     perpetuate
    1.73
     misog
    1.71
     authoritarian
    1.71
     perpet
    1.71
     capitalism
    1.69
     totalitarian
    1.67
     imperialism
    1.65
    Act Density 1.108%

    No Known Activations