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

    Neuronpedia

    INDEX
    Explanations

    occurrences of the word "sometimes."

    oai_token-act-pair · gpt-3.5-turbo

    the frequent use of the word "sometimes."

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    olitan
    -0.95
    mson
    -0.91
    zag
    -0.89
    rats
    -0.79
    ongyang
    -0.77
    ophy
    -0.75
    achus
    -0.74
    iem
    -0.74
    iak
    -0.74
    edIn
    -0.73
    POSITIVE LOGITS
     referred
    0.99
    hift
    0.89
    entimes
    0.89
     mistaken
    0.87
     even
    0.86
    theless
    0.82
     glimps
    0.81
     tempted
    0.81
     called
    0.81
     conflicting
    0.80
    Act Density 0.039%

    No Known Activations