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

    Neuronpedia

    INDEX
    Explanations

    Ratings/Numbers

    np_max-act · gemini-2.0-flash

    The neuron primarily detects numeric expressions—especially ratings, percentages, and viewer‐count figures.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    cou
    -0.08
    /ion
    -0.07
    �
    -0.06
    bp
    -0.06
    BP
    -0.06
     amort
    -0.06
     combines
    -0.06
     Soviets
    -0.06
     xấu
    -0.06
     담
    -0.06
    POSITIVE LOGITS
    (Syntax
    0.07
     MIC
    0.07
     widow
    0.07
    Purple
    0.06
    βε
    0.06
    -Bar
    0.06
    _NUM
    0.06
    	doc
    0.06
    عة
    0.06
    (Locale
    0.06
    Act Density 0.006%

    No Known Activations