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

    Neuronpedia

    INDEX
    Explanations

    proper nouns

    oai_token-act-pair · gpt-3.5-turbo

    sentences that contain punctuation, particularly periods, indicating the end of statements

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    itude
    -0.70
     sack
    -0.69
     bumped
    -0.67
    urance
    -0.66
     hug
    -0.66
     forgot
    -0.65
     forestry
    -0.65
     elbow
    -0.64
     consolation
    -0.63
     stash
    -0.62
    POSITIVE LOGITS
     Its
    1.41
     Originally
    1.20
     Unlike
    1.17
     Initially
    1.14
     Currently
    1.12
     It
    1.10
     Known
    1.09
     Although
    1.09
     Since
    1.08
     Typically
    1.06
    Act Density 0.384%

    No Known Activations