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

    Neuronpedia

    INDEX
    Explanations

    most common

    np_max-act · gemini-2.0-flash

    This neuron detects superlative-frequency phrases indicating the most common or most frequent items (e.g. “most common,” “most frequent,” “commonest,” “most prevalent”).

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
     sıras
    -0.07
    Negative
    -0.07
     humili
    -0.06
    анг
    -0.06
     bruises
    -0.06
    �
    -0.06
     вним
    -0.06
    OG
    -0.06
    umb
    -0.06
    كل
    -0.06
    POSITIVE LOGITS
     ア
    0.08
    [I
    0.07
    /library
    0.07
     []:↵
    0.07
    -cn
    0.06
                                ↵
    0.06
     ein
    0.06
     Afro
    0.06
    (I
    0.06
    吨
    0.06
    Act Density 0.024%

    No Known Activations