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

    Neuronpedia

    INDEX
    Explanations

    sub

    np_max-act · gemini-2.0-flash

    mathematical operations and their applications.

    oai_token-act-pair · gpt-4o-miniTriggered by @xinyanhu8

    The neuron selectively turns on for the Italian word “sottrazione” (i.e. any pieces of that subtraction‐related token).

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
     WHEN
    -0.07
    ática
    -0.07
     yurt
    -0.06
     justo
    -0.06
    rica
    -0.06
    　　　　　 　
    -0.06
     ($("#
    -0.06
     особенно
    -0.06
     wealth
    -0.06
    CompanyId
    -0.06
    POSITIVE LOGITS
    ignant
    0.07
    uggy
    0.06
    donnees
    0.06
     рес
    0.06
    exampleModalLabel
    0.06
     knowledgeable
    0.06
    -application
    0.06
    hy
    0.06
    ون
    0.06
     vak
    0.06
    Act Density 0.017%

    No Known Activations