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

    Neuronpedia

    INDEX
    Explanations

    math calculations

    np_max-act · gemini-2.0-flash

    The neuron fires on the shorthand “S” markers that introduce simplified‐result lines (i.e. the “S” tokens before the simplify outputs).

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    ,str
    -0.06
    fir
    -0.06
    _Char
    -0.06
     MAR
    -0.06
    /GL
    -0.06
    REL
    -0.06
    _LEN
    -0.06
    _RS
    -0.06
    %@",
    -0.06
    .Brand
    -0.06
    POSITIVE LOGITS
     Validator
    0.07
    Chicago
    0.07
    encia
    0.07
     города
    0.06
     isinstance
    0.06
     chtěl
    0.06
    uciones
    0.06
    只要
    0.06
    хови
    0.06
    Aliases
    0.06
    Act Density 0.016%

    No Known Activations