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

    Neuronpedia

    INDEX
    Explanations

    the word "the" at the start of sentences

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "the."

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    merce
    -0.75
    ocument
    -0.74
     accordingly
    -0.72
     furthermore
    -0.72
    Versions
    -0.71
     anew
    -0.68
    intosh
    -0.68
    Layer
    -0.65
    olson
    -0.64
     nevertheless
    -0.64
    POSITIVE LOGITS
     outset
    1.36
     standpoint
    1.23
     aforementioned
    1.10
     earliest
    0.96
     same
    0.96
     depths
    0.95
     perspective
    0.93
     confines
    0.91
     onset
    0.91
     smallest
    0.91
    Act Density 0.188%

    No Known Activations