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

    Neuronpedia

    INDEX
    Explanations

    the word "however"

    oai_token-act-pair · gpt-3.5-turbo

    the word "however."

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    enders
    -0.69
    uto
    -0.68
    jar
    -0.68
    ribes
    -0.67
    amed
    -0.66
    """
    -0.64
    mail
    -0.64
     Register
    -0.63
    lees
    -0.62
    mails
    -0.61
    POSITIVE LOGITS
     agre
    0.79
     interestingly
    0.79
    CLASSIFIED
    0.77
     guiActiveUn
    0.76
     srf
    0.76
    theless
    0.74
     reluct
    0.72
    querque
    0.71
     preferably
    0.70
    estyles
    0.69
    Act Density 0.016%

    No Known Activations