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

    Neuronpedia

    INDEX
    Explanations

    words related to citations, references, and acknowledgments

    oai_token-act-pair · gpt-3.5-turbo

    instances of punctuation, specifically commas, throughout the text

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    inar
    -0.76
    raq
    -0.68
    assed
    -0.66
    ose
    -0.66
    rors
    -0.65
    ¬¼
    -0.65
    Parameters
    -0.65
    icia
    -0.64
    emaker
    -0.64
    eman
    -0.63
    POSITIVE LOGITS
     however
    1.04
     there
    0.94
    there
    0.94
     although
    0.84
     "[
    0.82
     namely
    0.78
     though
    0.77
     we
    0.75
     THERE
    0.74
     quoting
    0.72
    Act Density 0.103%

    No Known Activations