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

    Neuronpedia

    INDEX
    Explanations

    PHP source code

    np_max-act · gemini-2.0-flash

    The neuron is detecting code identifier segments that contain underscores (e.g. parts of class or namespace names in PHP).

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    args
    -0.07
     interior
    -0.07
    rays
    -0.07
     مث
    -0.06
    Exp
    -0.06
    Prepare
    -0.06
    :def
    -0.06
    (cost
    -0.06
     cs
    -0.06
    旅
    -0.06
    POSITIVE LOGITS
     Ricardo
    0.07
     reopening
    0.06
     rune
    0.06
    "]=>
    0.06
    '</
    0.06
     kaynağı
    0.06
    ("/
    0.06
     moderation
    0.06
    uncan
    0.06
     Ping
    0.06
    Act Density 0.004%

    No Known Activations