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

    Neuronpedia

    INDEX
    Explanations

    phrases related to commentary or news items

    oai_token-act-pair · gpt-3.5-turbo
    New Auto-Interp
    Negative Logits
     harbor
    -0.72
     Icelandic
    -0.63
     Aval
    -0.63
     spoil
    -0.62
     bitters
    -0.61
     Shining
    -0.61
     encl
    -0.60
     recoil
    -0.60
     Mara
    -0.58
     Hayden
    -0.58
    POSITIVE LOGITS
    fortable
    1.47
    pleting
    1.47
    plement
    1.38
    puters
    1.38
    pletion
    1.37
    plex
    1.35
    ptroller
    1.31
    mented
    1.27
    puting
    1.26
    rade
    1.26
    Act Density 0.014%

    No Known Activations