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

    Neuronpedia

    INDEX
    Explanations

    contractions with an apostrophe or specific stylistic choices like ending sentences with a period and a comma

    oai_token-act-pair · gpt-3.5-turbo

    phrases indicating emotional states or feelings

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    inki
    -0.85
    ê
    -0.78
    ée
    -0.76
    Bah
    -0.74
    waukee
    -0.74
     Kelvin
    -0.74
    EVA
    -0.73
    ë
    -0.72
    elin
    -0.71
    ÃŃn
    -0.70
    POSITIVE LOGITS
     Sc
    2.53
    Sc
    2.31
    sc
    2.05
     sc
    1.88
     SC
    1.86
    SC
    1.67
     Scar
    1.33
     Scrib
    1.32
     Scal
    1.31
     Scorp
    1.30
    Act Density 0.210%

    No Known Activations