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

    Neuronpedia

    INDEX
    Explanations

    phrases instructing or explaining a simple process

    oai_token-act-pair · gpt-3.5-turbo

    the word "Simply" and its variations used for giving instructions or explanations

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
     elders
    -0.70
     attendant
    -0.70
     coun
    -0.66
     extensively
    -0.64
     patent
    -0.64
     protector
    -0.64
     vigorously
    -0.63
    egal
    -0.62
     meticulously
    -0.58
     particularly
    -0.58
    POSITIVE LOGITS
    IFIED
    0.87
    tons
    0.86
    Simply
    0.85
    ername
    0.80
    reet
    0.76
    Login
    0.75
    Drag
    0.75
    ified
    0.75
    isine
    0.74
    reme
    0.73
    Act Density 0.011%

    No Known Activations