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

    Neuronpedia

    INDEX
    Explanations

    phrases or sentences enclosed in quotation marks

    oai_token-act-pair · gpt-3.5-turbo

    quoted statements or direct speech within the text

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
     vault
    -0.73
     prelim
    -0.68
     arch
    -0.68
     upstairs
    -0.67
     downstairs
    -0.67
     casc
    -0.66
     spine
    -0.66
     footing
    -0.65
     sprint
    -0.65
     crunch
    -0.65
    POSITIVE LOGITS
    We
    1.40
    Our
    1.30
    There
    1.30
    People
    1.29
    It
    1.26
    Because
    1.24
    Today
    1.23
    Certainly
    1.23
    Unless
    1.22
    They
    1.22
    Act Density 0.108%

    No Known Activations