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

    Neuronpedia

    INDEX
    Explanations

    asterisk

    np_max-act · gemini-2.0-flash

    The neuron selectively activates on high-precision decimal numbers (e.g. “1.2802734375,” “0.5517578125,” etc.), i.e. floating-point numeric tokens.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    _dy
    -0.07
    ราะ
    -0.06
    _VF
    -0.06
    言って
    -0.06
    170
    -0.06
     fractions
    -0.06
     wk
    -0.06
     rigged
    -0.06
    iễ
    -0.06
     EPA
    -0.06
    POSITIVE LOGITS
     afirm
    0.07
     Mass
    0.07
     Nations
    0.07
    .addItem
    0.07
     equipped
    0.07
     nás
    0.06
    ivos
    0.06
     النس
    0.06
     philanth
    0.06
    Esta
    0.06
    Act Density 0.002%

    No Known Activations