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

    Neuronpedia

    INDEX
    Explanations

    phrases related to ordering, coordination, and instructions

    oai_token-act-pair · gpt-3.5-turbo

    punctuation, particularly sentence-ending periods

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
     instinct
    -0.82
    ŃĶ
    -0.79
    emic
    -0.76
     tangled
    -0.75
     buried
    -0.73
     footing
    -0.73
     bashing
    -0.73
     tackling
    -0.71
     shedding
    -0.71
     ugly
    -0.71
    POSITIVE LOGITS
     Additionally
    1.49
     Alternatively
    1.45
     Therefore
    1.38
     Please
    1.37
     However
    1.36
     Otherwise
    1.34
     Depending
    1.33
     Furthermore
    1.26
     Also
    1.24
     Likewise
    1.23
    Act Density 0.474%

    No Known Activations