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

    Neuronpedia

    INDEX
    Explanations

    negative or alarming terms

    oai_token-act-pair · gpt-3.5-turbo

    terms indicating negative attributes or serious concerns

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
     Rath
    -0.51
     Quan
    -0.49
     flight
    -0.48
     pup
    -0.48
     bursting
    -0.47
     booted
    -0.45
     saturated
    -0.44
     ex
    -0.44
     fertile
    -0.44
     brisk
    -0.43
    POSITIVE LOGITS
    terday
    1.03
    theless
    0.95
    etheless
    0.83
    tenance
    0.83
    withstanding
    0.82
    lihood
    0.79
    veyard
    0.79
    mosp
    0.78
    odore
    0.75
    ãĥĥ
    0.72
    Act Density 0.659%

    No Known Activations