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

    Neuronpedia

    INDEX
    Explanations

    dash

    np_max-act · gemini-2.0-flash

    The neuron fires on the hyphen-separated metadata field names in gettext PO file headers (e.g. keys like “POT-Creation-Date,” “PO-Revision-Date,” “Last-Translator,” etc.).

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    inx
    -0.07
     Mend
    -0.06
    eb
    -0.06
     buffered
    -0.06
    朗
    -0.06
     Rice
    -0.06
    Drawer
    -0.06
    \Builder
    -0.06
     Huck
    -0.06
     accusations
    -0.06
    POSITIVE LOGITS
     onUpdate
    0.07
     CASCADE
    0.06
    hPa
    0.06
    setTitle
    0.06
     Yosemite
    0.06
    飛
    0.06
    生产
    0.06
    GetX
    0.06
     Investing
    0.06
     graphql
    0.06
    Act Density 0.001%

    No Known Activations