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

    Neuronpedia

    INDEX
    Explanations

    a

    np_max-act · gemini-2.0-flash

    This neuron detects when the text is describing marking or annotating a method or class (e.g. statements like “mark a method or a class as …”).

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
     гром
    -0.07
     Election
    -0.07
    Token
    -0.06
    _arc
    -0.06
     bran
    -0.06
     Hep
    -0.06
     ton
    -0.06
     election
    -0.06
    sf
    -0.06
     gens
    -0.06
    POSITIVE LOGITS
     assured
    0.08
     els
    0.07
    .,↵
    0.07
    estado
    0.07
    .cz
    0.07
     mně
    0.06
    _UNSIGNED
    0.06
     tedy
    0.06
     tostring
    0.06
    _validate
    0.06
    Act Density 0.063%

    No Known Activations