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

    Neuronpedia

    INDEX
    Explanations

    quotations

    oai_token-act-pair · gpt-3.5-turbo

    quotation marks indicating direct speech or dialogue

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
     appro
    -0.79
     disappro
    -0.78
     lowly
    -0.77
     favour
    -0.77
     favor
    -0.72
     departed
    -0.71
     scrimmage
    -0.70
     discont
    -0.69
     pir
    -0.68
     insider
    -0.67
    POSITIVE LOGITS
    We
    1.44
    They
    1.39
    There
    1.38
    It
    1.37
    Our
    1.34
    Basically
    1.33
    Everybody
    1.32
    Certainly
    1.31
    Nobody
    1.31
    People
    1.30
    Act Density 0.123%

    No Known Activations