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

    Neuronpedia

    INDEX
    Explanations

    phrases related to challenges or obstacles

    oai_token-act-pair · gpt-3.5-turbo

    punctuation and sentence structure

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    but
    -1.12
     but
    -1.05
    butt
    -0.79
    odi
    -0.76
     BUT
    -0.72
    However
    -0.69
    But
    -0.68
    BUT
    -0.65
    hex
    -0.61
    istant
    -0.61
    POSITIVE LOGITS
     nonetheless
    0.88
    etheless
    0.80
    answered
    0.73
     Says
    0.69
     Moreover
    0.67
     Meaning
    0.66
     Therefore
    0.65
    *****
    0.64
     Hence
    0.63
     Worse
    0.61
    Act Density 0.774%

    No Known Activations