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

    Neuronpedia

    INDEX
    Explanations

    directions

    np_max-act · gemini-2.0-flash

    The neuron is detecting street addresses and location indicators (numbers, directional abbreviations, and block/exits) in text.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
     fib
    -0.07
     Wheels
    -0.07
    father
    -0.07
    Pages
    -0.07
     branded
    -0.07
    rollback
    -0.07
     lanc
    -0.06
    ач
    -0.06
    cki
    -0.06
     Codec
    -0.06
    POSITIVE LOGITS
     procrast
    0.08
    /dr
    0.07
     سرم
    0.07
    (dec
    0.07
     کردن
    0.07
    .art
    0.06
    ″N
    0.06
     може
    0.06
     LES
    0.06
    音樂
    0.06
    Act Density 0.003%

    No Known Activations