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

    Neuronpedia

    INDEX
    Explanations

    physical harm or danger related words and situations

    oai_token-act-pair · gpt-3.5-turbo
    New Auto-Interp
    Negative Logits
    omorph
    -0.26
    ioxide
    -0.24
    ramid
    -0.23
    oter
    -0.23
    aminer
    -0.22
    alogy
    -0.22
    arta
    -0.21
    etheus
    -0.21
    othermal
    -0.21
    anyon
    -0.21
    POSITIVE LOGITS
    enance
    0.27
    iate
    0.27
    soever
    0.25
    Compat
    0.25
    adoes
    0.25
    yet
    0.24
    theless
    0.24
    withstanding
    0.23
     whatsoever
    0.23
    iless
    0.23
    Act Density 15.257%

    No Known Activations