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

    Neuronpedia

    INDEX
    Explanations

    phrases or words related to strong disapproval or criticism

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "condemned."

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    thora
    -0.93
    ovember
    -0.79
    vati
    -0.78
    weeney
    -0.76
    opic
    -0.73
    ramid
    -0.71
    athering
    -0.70
    hack
    -0.70
    OTE
    -0.69
    nown
    -0.69
    POSITIVE LOGITS
     condemned
    0.92
     condemning
    0.82
     condemn
    0.82
     condemns
    0.79
     harshly
    0.78
    utsche
    0.77
     condem
    0.75
     criticised
    0.69
     critic
    0.68
     Survivors
    0.68
    Act Density 0.009%

    No Known Activations