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

    Neuronpedia

    INDEX
    Explanations

    -page

    np_max-act · gemini-2.0-flash

    The neuron detects numeric page-number tokens (floating-point values) that follow “scanned-page.”

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
     eos
    -0.06
     ')↵↵
    -0.06
    كام
    -0.06
     stirring
    -0.06
     Bez
    -0.06
     entitled
    -0.06
    !↵↵
    -0.06
     mon
    -0.05
    ));↵↵
    -0.05
    点
    -0.05
    POSITIVE LOGITS
     dang
    0.07
     лікар
    0.07
    ABB
    0.07
    .Ph
    0.07
     üzerine
    0.07
     cultiv
    0.07
     Promo
    0.07
    っち
    0.06
    atcher
    0.06
    	meta
    0.06
    Act Density 0.001%

    No Known Activations