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

    Neuronpedia

    INDEX
    Explanations

    _H

    np_max-act · gemini-2.0-flash

    The neuron is detecting C/C++ preprocessor directive tokens (e.g. #ifdef, #include, #endif).

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    -policy
    -0.08
     ліка
    -0.07
     entrepreneurs
    -0.07
    чає
    -0.07
     Poh
    -0.06
    .border
    -0.06
     Ach
    -0.06
    GLenum
    -0.06
    their
    -0.06
    scal
    -0.06
    POSITIVE LOGITS
     ‎
    0.06
     lucky
    0.06
     explores
    0.06
     alongside
    0.06
     banc
    0.06
    uencia
    0.05
    (nullptr
    0.05
     drew
    0.05
    east
    0.05
     поврежд
    0.05
    Act Density 0.002%

    No Known Activations