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

    Neuronpedia

    INDEX
    Explanations

    phrases related to turning on and off electronic devices or lights

    oai_token-act-pair · gpt-3.5-turbo

    phrases related to turning things on and off

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    ijk
    -0.84
    gerald
    -0.80
    bris
    -0.78
    owicz
    -0.78
    ylum
    -0.76
    legate
    -0.75
    bourg
    -0.75
    ossal
    -0.68
    iago
    -0.67
    obos
    -0.66
    POSITIVE LOGITS
     ignition
    1.38
     headlights
    1.27
     lights
    1.25
     LEDs
    1.12
     ther
    1.11
     alarms
    1.09
     brakes
    1.07
     switches
    1.07
     wip
    1.05
     heater
    1.05
    Act Density 0.209%

    No Known Activations