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

    Neuronpedia

    INDEX
    Explanations

    expressions involving alternative methods or perspectives

    oai_token-act-pair · gpt-3.5-turbo

    instances where the phrase "as a way" is used to explain purpose or intention

    oai_token-act-pair · gpt-4o-miniTriggered by @bot
    New Auto-Interp
    Negative Logits
    ancies
    -0.81
    avorite
    -0.73
    asions
    -0.71
     Sounds
    -0.62
    utenberg
    -0.62
    eds
    -0.62
    inately
    -0.62
    urations
    -0.61
    cffff
    -0.61
    ortunately
    -0.60
    POSITIVE LOGITS
    ward
    0.98
    point
    0.93
    fare
    0.93
    forward
    0.83
     forward
    0.82
    finding
    0.80
    ÙĨ
    0.75
    station
    0.74
    points
    0.72
    WARD
    0.71
    Act Density 0.082%

    No Known Activations