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

    Neuronpedia

    INDEX
    Explanations

    words related to illegal activities, especially those related to deception or coercion, like blackmail and extortion

    oai_token-act-pair · gpt-3.5-turbo

    words and phrases related to criminal activities, particularly extortion and blackmail

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    çĦ
    -0.82
    =-=-=-=-=-=-=-=-
    -0.76
     McA
    -0.73
     Mood
    -0.72
    ixed
    -0.72
    ________________
    -0.68
    aug
    -0.68
     Atmospheric
    -0.67
    acht
    -0.67
    Frames
    -0.67
    POSITIVE LOGITS
     blackmail
    1.25
     extortion
    1.22
     challeng
    0.86
     kidnapping
    0.80
     coercion
    0.80
    ate
    0.78
     sorcery
    0.76
     robbery
    0.75
     stalking
    0.73
     solicitation
    0.73
    Act Density 0.018%

    No Known Activations