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

    Neuronpedia

    INDEX
    Explanations

    code and programming

    np_max-act · gemini-2.0-flash

    This neuron fires on tokens that are part of angle-bracket syntax—i.e. the “<...>” portions of markup tags or include directives.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    hexdigest
    -0.07
     Razor
    -0.06
    	l
    -0.06
    .encrypt
    -0.06
    .Tools
    -0.06
     Method
    -0.06
    ραν
    -0.06
    -process
    -0.06
     deviation
    -0.06
    _capability
    -0.06
    POSITIVE LOGITS
     çocu
    0.07
    "(
    0.06
     υ
    0.06
     punitive
    0.06
    Sadly
    0.06
     Sadly
    0.06
    .tencent
    0.06
    เว
    0.06
     lucr
    0.06
     Freund
    0.06
    Act Density 0.032%

    No Known Activations