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

    Neuronpedia

    INDEX
    Explanations

    contractions of "have not"

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "haven't."

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
     princ
    -0.71
     flies
    -0.64
     impressions
    -0.59
     treadmill
    -0.58
     manual
    -0.58
     simpl
    -0.58
     fly
    -0.57
     couch
    -0.57
     descending
    -0.56
     Cellular
    -0.56
    POSITIVE LOGITS
    't
    1.63
    ÃŃ
    1.14
    ited
    1.09
    ´
    0.91
    ãĤ§
    0.90
    dayName
    0.86
    iting
    0.86
    kered
    0.85
    its
    0.84
    itely
    0.83
    Act Density 0.048%

    No Known Activations