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

    Neuronpedia

    INDEX
    Explanations

    phrases or words related to correcting something

    oai_token-act-pair · gpt-3.5-turbo

    variations of the word "correct."

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    atos
    -0.97
     Pry
    -0.72
    CHO
    -0.72
    nam
    -0.72
    thin
    -0.71
    atten
    -0.71
     Valhalla
    -0.71
    bang
    -0.70
    NetMessage
    -0.69
    aden
    -0.66
    POSITIVE LOGITS
     spelling
    1.00
     misinformation
    0.94
     inaccur
    0.84
     erroneous
    0.83
    able
    0.81
     pronunciation
    0.79
    ibly
    0.76
    fully
    0.74
    ions
    0.74
    orate
    0.74
    Act Density 0.042%

    No Known Activations