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

    Neuronpedia

    INDEX
    Explanations

    words related to contrast or alternative choices

    oai_token-act-pair · gpt-3.5-turbo

    expressions of alternatives or contrasts

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    eele
    -0.77
    owe
    -0.72
    âĹ¼
    -0.69
    eno
    -0.62
    CLUD
    -0.62
     Influ
    -0.62
    çĶ°
    -0.61
    acha
    -0.61
    quad
    -0.61
    oker
    -0.60
    POSITIVE LOGITS
     blindly
    1.02
     altogether
    0.98
     passively
    0.94
     outright
    0.87
     anymore
    0.86
     solely
    0.83
     ourselves
    0.81
     entirely
    0.81
     mindless
    0.77
     oneself
    0.76
    Act Density 0.256%

    No Known Activations