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

    Neuronpedia

    INDEX
    Explanations

    phrases indicating downward movement or negative progression

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    bian
    -0.17
    ione
    -0.16
     up
    -0.16
    aurant
    -0.16
    inness
    -0.16
    ж
    -0.15
    quette
    -0.15
    itag
    -0.15
    ynom
    -0.14
    kate
    -0.14
    POSITIVE LOGITS
    graded
    0.24
    wards
    0.23
    /up
    0.21
    grades
    0.20
    played
    0.19
    /down
    0.19
    grading
    0.19
    stairs
    0.19
    WARDS
    0.18
    beat
    0.18
    Act Density 0.056%

    No Known Activations