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

    Neuronpedia

    INDEX
    Explanations

    news agency names in captions

    oai_token-act-pair · gpt-3.5-turbo

    instances of the word "REUTERS" in the text

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
     sed
    -0.69
     intellig
    -0.58
     tranquil
    -0.57
    bal
    -0.56
     tire
    -0.56
    inent
    -0.56
     bra
    -0.56
    izers
    -0.56
     perplex
    -0.55
     discipl
    -0.55
    POSITIVE LOGITS
    UTERS
    0.90
     Images
    0.86
    CLASSIFIED
    0.85
     PHOTO
    0.80
    photo
    0.78
    OGR
    0.78
    FORE
    0.78
    Film
    0.77
     IMAGES
    0.77
    PLE
    0.77
    Act Density 0.017%

    No Known Activations