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

    Neuronpedia

    INDEX
    Explanations

    short phrases followed by intensifiers or modifiers like 'bit', 'pretty', 'good', 'whole', etc

    oai_token-act-pair · gpt-3.5-turbo

    expressions of quantity and assessment

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    hours
    -0.79
    effects
    -0.77
    anqu
    -0.75
    bots
    -0.70
    alks
    -0.69
    Eng
    -0.69
    asketball
    -0.68
    Autom
    -0.68
    today
    -0.67
    dule
    -0.67
    POSITIVE LOGITS
     typo
    1.09
     lot
    1.09
     recipe
    1.06
     misconception
    1.06
     mistake
    1.04
     understatement
    1.03
     pretty
    1.02
     contradiction
    1.02
     prerequisite
    1.01
     rarity
    1.01
    Act Density 0.157%

    No Known Activations