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

    Neuronpedia

    INDEX
    Explanations

    gibberish text

    np_max-act · gemini-2.0-flash

    The neuron responds to short non‐word tokens—especially sequences of uppercase letters, acronyms, or symbol‐heavy codes.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    914
    -0.07
     důležit
    -0.07
     merry
    -0.06
    	engine
    -0.06
    subscriber
    -0.06
    atrigesimal
    -0.06
    kus
    -0.06
     спів
    -0.06
     ****************************************
    -0.06
    cidade
    -0.06
    POSITIVE LOGITS
    .TrimSpace
    0.07
    ，由
    0.06
     nomin
    0.06
    oen
    0.06
    .onNext
    0.06
    loe
    0.06
    lop
    0.06
    .Fatal
    0.06
     amazon
    0.06
    ‚Ì
    0.06
    Act Density 0.009%

    No Known Activations