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

    Neuronpedia

    INDEX
    Explanations

    Security vulnerabilities

    np_max-act · gemini-2.0-flash

    The neuron is activated by numeric literal tokens (especially floating-point numbers) in the text.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    isAdmin
    -0.06
    Injected
    -0.06
     objectMapper
    -0.06
    Emoji
    -0.06
     Islands
    -0.06
    _validation
    -0.06
    .GPIO
    -0.06
    adol
    -0.06
    izioni
    -0.06
    انس
    -0.06
    POSITIVE LOGITS
    Living
    0.07
     liv
    0.06
    /********************************
    0.06
     آور
    0.06
     derives
    0.06
     kork
    0.06
     삼성
    0.06
    	UPROPERTY
    0.06
     their
    0.06
     masturb
    0.06
    Act Density 0.027%

    No Known Activations