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

    Neuronpedia

    INDEX
    Explanations

    instances where the text discusses specific conditions or situations relevant to the reader

    oai_token-act-pair · gpt-3.5-turbo

    the word "and" along with various conditions and scenarios involving the reader

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    Contents
    -0.68
    ãĥ³ãĤ¸
    -0.63
    DIV
    -0.60
    rawdownloadcloneembedreportprint
    -0.60
    luster
    -0.59
    Bind
    -0.59
    forth
    -0.57
    arious
    -0.57
    ãĤ´ãĥ³
    -0.57
     Flavoring
    -0.56
    POSITIVE LOGITS
     want
    1.11
     aren
    1.08
    want
    1.02
     weren
    0.98
     haven
    0.92
     decide
    0.91
     are
    0.89
     wanna
    0.87
     intend
    0.86
     wish
    0.86
    Act Density 0.164%

    No Known Activations