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

    Neuronpedia

    INDEX
    Explanations

    far

    np_max-act · gemini-2.0-flash

    This neuron detects degree-of-comparison and intensifier phrases, especially superlative emphasis like “by far the most.”

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
     Kanunu
    -0.08
    .getZ
    -0.07
    公開
    -0.07
     Nullable
    -0.07
    दर
    -0.07
    (renderer
    -0.07
     Persistence
    -0.06
    奶
    -0.06
    Popover
    -0.06
    (yy
    -0.06
    POSITIVE LOGITS
     far
    0.07
    тра
    0.06
    ále
    0.06
     denim
    0.06
    _raise
    0.06
    登
    0.06
     fanatic
    0.06
     closet
    0.06
    ายใน
    0.06
    astos
    0.06
    Act Density 0.036%

    No Known Activations