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

    Neuronpedia

    INDEX
    Explanations

    phrases indicating a quantity or range, such as "three or four."

    oai_token-act-pair · gpt-3.5-turbo

    references to quantities, particularly the numbers three and four, indicating frequency or count

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    /#
    -0.72
    lance
    -0.72
    trust
    -0.67
     Devi
    -0.66
    laws
    -0.66
    Stat
    -0.65
    nah
    -0.64
    Joy
    -0.63
    Official
    -0.62
    ivism
    -0.62
    POSITIVE LOGITS
     thousand
    0.81
     miles
    0.78
     mph
    0.75
     percent
    0.74
     hundred
    0.74
    undred
    0.73
     sectors
    0.73
     gigs
    0.73
     minutes
    0.72
     strokes
    0.71
    Act Density 0.084%

    No Known Activations