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

    Neuronpedia

    INDEX
    Explanations

    mentions of actions done gently or with little force

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "lightly" and its variations

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    aura
    -0.72
    ceiver
    -0.70
    pit
    -0.69
    tv
    -0.65
    earances
    -0.65
    global
    -0.62
    cat
    -0.62
    tu
    -0.62
    Bern
    -0.61
    history
    -0.61
    POSITIVE LOGITS
     lightly
    4.01
     softly
    1.69
     gently
    1.64
     loosely
    1.59
     casually
    1.38
     hastily
    1.35
     cautiously
    1.33
     mildly
    1.32
     generously
    1.32
     heavily
    1.31
    Act Density 0.014%

    No Known Activations