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

    Neuronpedia

    INDEX
    Explanations

    references or comparisons

    np_max-act · gemini-2.0-flash

    This neuron activates on tokens representing precise numeric values—especially floating-point decimals.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    uluk
    -0.07
     тис
    -0.06
     пл
    -0.06
    Tak
    -0.06
     Kürt
    -0.06
     Кор
    -0.06
     ổ
    -0.06
    oltage
    -0.06
    Tue
    -0.06
    vang
    -0.06
    POSITIVE LOGITS
     recounted
    0.07
    ................................................................
    0.06
    .Write
    0.06
    }")]↵
    0.06
    ')."
    0.06
    ................................................................
    0.06
     Frm
    0.06
    .Raise
    0.06
    (hw
    0.06
     dinheiro
    0.06
    Act Density 0.148%

    No Known Activations