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 fires on lines of code or markup syntax—identifying code‐block tokens like tags, braces, and CSS/JS constructs.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    .user
    -0.06
    _threads
    -0.06
    Metric
    -0.06
     est
    -0.06
    -ves
    -0.06
    두
    -0.06
    ilmek
    -0.06
     <<-
    -0.06
     rend
    -0.06
    .bits
    -0.06
    POSITIVE LOGITS
    ivo
    0.07
     accordance
    0.06
    нят
    0.06
     Lift
    0.06
    IVO
    0.06
     jedná
    0.06
     خانه
    0.06
     REQUIRE
    0.06
    　　　　 　 　
    0.06
     máme
    0.06
    Act Density 0.003%

    No Known Activations