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

    Neuronpedia

    INDEX
    Explanations

    obj

    np_max-act · gemini-2.0-flash

    This neuron detects C++ template parameter delimiters (the angle-bracketed `<…>` sequences in template/generic declarations).

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    (factor
    -0.07
    rav
    -0.07
     forts
    -0.07
     Could
    -0.06
    haled
    -0.06
    ndo
    -0.06
    erer
    -0.06
    ต
    -0.06
     ilet
    -0.06
    医疗
    -0.06
    POSITIVE LOGITS
     составе
    0.06
     clazz
    0.06
    .att
    0.06
     Inside
    0.06
    ${
    0.05
     Dhabi
    0.05
     Visitors
    0.05
    撃
    0.05
     ofere
    0.05
     Appe
    0.05
    Act Density 0.010%

    No Known Activations