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

    Neuronpedia

    INDEX
    Explanations

    instances where there is room for improvement or development

    oai_token-act-pair · gpt-3.5-turbo

    phrases indicating potential for improvement or additional opportunities

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    é¾
    -0.83
    loo
    -0.67
    hz
    -0.67
    gypt
    -0.64
    cles
    -0.63
    geist
    -0.63
    owsky
    -0.62
     averaged
    -0.62
    cz
    -0.62
    nick
    -0.62
    POSITIVE LOGITS
    bidden
    0.94
     accommodation
    0.87
     starters
    0.87
    gotten
    0.85
     hire
    0.84
     storing
    0.82
     inclusion
    0.80
     example
    0.79
     conting
    0.76
     purposes
    0.76
    Act Density 0.152%

    No Known Activations