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

    Neuronpedia

    INDEX
    Explanations

    legal/financial documents

    np_max-act · gemini-2.0-flash

    This neuron activates on numeric tokens—especially legal citation numbers (e.g. article, section, paragraph, and other numeric references) in the text.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
     схем
    -0.07
    聯
    -0.06
     ).↵
    -0.06
     coordin
    -0.06
     ),
    -0.06
     škol
    -0.06
    -0.06
     Enterprises
    -0.06
    ).↵
    -0.06
    Thanks
    -0.06
    POSITIVE LOGITS
    >'
    0.07
     boxes
    0.07
     Pdf
    0.06
     beforeSend
    0.06
     seize
    0.06
    otech
    0.06
     occurring
    0.06
    >Delete
    0.06
     realloc
    0.06
     OTHER
    0.06
    Act Density 0.003%

    No Known Activations