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

    Neuronpedia

    INDEX
    Explanations

    quantifiers

    np_max-act · gemini-2.0-flash

    The neuron selectively activates on numeric tokens, especially decimal numbers and digit‐containing measurements.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    ictionary
    -0.07
     Orr
    -0.07
     making
    -0.07
     Lopez
    -0.06
    quiz
    -0.06
     Order
    -0.06
     Kong
    -0.06
     widening
    -0.06
     Hernandez
    -0.06
    --------↵↵
    -0.06
    POSITIVE LOGITS
    aison
    0.08
    statt
    0.07
    스가
    0.06
     далі
    0.06
    Ģ
    0.06
     "()
    0.06
    Tôi
    0.06
     zwar
    0.06
     مستقیم
    0.06
     tends
    0.06
    Act Density 0.070%

    No Known Activations