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

    Neuronpedia

    INDEX
    Explanations

    phrases related to an argument or justification, usually starting with "For."

    oai_token-act-pair · gpt-3.5-turbo

    the phrase "For [number]" indicating promotions or offers

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
     close
    -0.69
     thought
    -0.68
     sustained
    -0.64
     rank
    -0.64
     causation
    -0.63
     near
    -0.63
     dent
    -0.63
     apparently
    -0.60
     sounding
    -0.59
     breathing
    -0.59
    POSITIVE LOGITS
    For
    2.80
    To
    1.81
    With
    1.72
     For
    1.71
    FOR
    1.66
    If
    1.63
    From
    1.62
    Against
    1.60
    for
    1.58
    Because
    1.58
    Act Density 0.022%

    No Known Activations