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

    Neuronpedia

    INDEX
    Explanations

    identifiers in code

    np_max-act · gemini-2.0-flash

    This neuron activates on references to field or column identifiers—especially tokens like “ID,” “name,” “key,” and similar schema/attribute labels.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    _OWNER
    -0.07
     vidéo
    -0.06
    iOS
    -0.06
    argest
    -0.06
     intercepted
    -0.06
    _MINOR
    -0.06
    _Success
    -0.06
    broker
    -0.06
     safest
    -0.06
    ")}
    -0.06
    POSITIVE LOGITS
     happy
    0.07
     Bright
    0.07
     regularly
    0.06
    ディ
    0.06
     Doctors
    0.06
     festival
    0.06
     <=
    0.06
     feminism
    0.06
     setIs
    0.06
     celebration
    0.06
    Act Density 0.063%

    No Known Activations