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

    Neuronpedia

    INDEX
    Explanations

    phrases related to actions or instructions involving the reader

    oai_token-act-pair · gpt-3.5-turbo

    phrases that directly address or involve the reader

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    course
    -0.68
     Sunset
    -0.66
     Meadow
    -0.65
    İĭ
    -0.65
    stadt
    -0.63
     Reconstruction
    -0.63
    duty
    -0.62
    odor
    -0.61
    =~
    -0.59
    Mission
    -0.58
    POSITIVE LOGITS
    're
    1.51
    've
    1.25
     hear
    0.98
     guys
    0.93
     realize
    0.90
    'll
    0.90
     compare
    0.88
     realise
    0.87
    'd
    0.86
     ask
    0.85
    Act Density 0.093%

    No Known Activations