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

    Neuronpedia

    INDEX
    Explanations

    math problems

    np_max-act · gemini-2.0-flash

    The neuron fires on numeric literal tokens that are floating‐point decimals (i.e. numbers expressed with a decimal point and many fractional digits).

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    pip
    -0.07
     Evangel
    -0.07
     architect
    -0.07
    ForeignKey
    -0.07
    /fr
    -0.07
    .easing
    -0.07
    _AM
    -0.07
    .Side
    -0.07
     Made
    -0.06
    Facebook
    -0.06
    POSITIVE LOGITS
    �
    0.07
    ynn
    0.06
     Garrison
    0.06
     comprehensive
    0.06
    строй
    0.06
    �
    0.06
     člen
    0.06
     breat
    0.06
     вниз
    0.06
     Cres
    0.06
    Act Density 0.002%

    No Known Activations