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

    Neuronpedia

    INDEX
    Explanations

    words related to light sources like candles and lanterns

    oai_token-act-pair · gpt-3.5-turbo

    references to candles and related light sources

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    hog
    -0.85
    rule
    -0.72
    FactoryReloaded
    -0.72
    ocard
    -0.71
    amaru
    -0.70
    hod
    -0.70
    icle
    -0.69
    ãĥį
    -0.68
    icles
    -0.66
    ramid
    -0.65
    POSITIVE LOGITS
     Lantern
    0.79
    ENSE
    0.76
     stove
    0.72
    atives
    0.72
    ENDED
    0.72
    lit
    0.69
     flame
    0.68
     cord
    0.67
    perty
    0.66
    belt
    0.65
    Act Density 0.091%

    No Known Activations