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

    Neuronpedia

    INDEX
    Explanations

    instances where something reaches an extreme or critical level

    oai_token-act-pair · gpt-3.5-turbo

    phrases indicating a significant or critical threshold

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    uthor
    -0.74
    annis
    -0.73
     Carbuncle
    -0.67
    ccording
    -0.67
    avorite
    -0.64
    omers
    -0.64
     notor
    -0.63
    æ©
    -0.63
    ãģķ
    -0.61
    ÄŁ
    -0.61
    POSITIVE LOGITS
    lessness
    1.07
    lessly
    0.93
    ioned
    0.80
    ifications
    0.80
    acle
    0.79
    osphere
    0.78
    posts
    0.78
    less
    0.77
    points
    0.75
    point
    0.75
    Act Density 0.021%

    No Known Activations