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

    Neuronpedia

    INDEX
    Explanations

    phrases related to tidying up or cleaning

    oai_token-act-pair · gpt-3.5-turbo

    phrases that include the word "up."

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    iw
    -0.66
    oth
    -0.62
    ayer
    -0.60
    oway
    -0.58
     Walters
    -0.58
    ow
    -0.58
    gemony
    -0.57
    Zero
    -0.57
    leans
    -0.57
    ====
    -0.57
    POSITIVE LOGITS
    dates
    0.90
    river
    0.82
    stairs
    0.71
    dating
    0.70
    grading
    0.69
    rights
    0.68
     shop
    0.68
    adesh
    0.67
    grades
    0.66
    raised
    0.64
    Act Density 0.095%

    No Known Activations