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 programming‐language syntax elements—keywords, operators, or tags (e.g. Java’s “new,” CSS’s “!important,” HTML tags like “<th>”/“<title>,” and similar code tokens).

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
     settled
    -0.07
     patents
    -0.07
    +"
    -0.07
    (total
    -0.07
    localized
    -0.07
    /config
    -0.07
    .rawQuery
    -0.07
    ϊ
    -0.07
    Installer
    -0.07
    şa
    -0.07
    POSITIVE LOGITS
    rganization
    0.07
    ,
    0.06
     ZX
    0.06
     taj
    0.06
    TECT
    0.06
     EF
    0.06
    Concurrency
    0.06
     подт
    0.06
     kıl
    0.06
     TREE
    0.05
    Act Density 0.052%

    No Known Activations