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

    Neuronpedia

    INDEX
    Explanations

    code

    np_max-act · gemini-2.0-flash

    This neuron detects code declaration keywords and types (e.g., “public,” “int,” “void,” identifiers following them) in programming-language source.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
     Bob
    -0.07
    _version
    -0.07
    (Book
    -0.07
     gốc
    -0.07
    .Font
    -0.07
    	Output
    -0.07
     Scheduler
    -0.07
     Liu
    -0.07
    ạo
    -0.07
     joyful
    -0.07
    POSITIVE LOGITS
    ght
    0.07
    ‍
    0.07
     odense
    0.06
    maze
    0.06
    야
    0.06
     영상
    0.06
     청
    0.06
    $product
    0.06
     меди
    0.06
    �
    0.06
    Act Density 0.021%

    No Known Activations