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

    Neuronpedia

    INDEX
    Explanations

    phrases related to errors or mistakes

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "error" and related terms

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    apeake
    -0.83
    apy
    -0.82
    amen
    -0.81
    tsky
    -0.81
    electric
    -0.80
    nai
    -0.80
    bledon
    -0.78
    estine
    -0.78
    Electric
    -0.76
    APTER
    -0.71
    POSITIVE LOGITS
     error
    0.91
     guiActiveUn
    0.88
    gered
    0.87
     margin
    0.84
     errors
    0.84
    ously
    0.82
     dece
    0.79
     mishand
    0.78
     deceive
    0.77
     message
    0.74
    Act Density 0.022%

    No Known Activations