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

    Neuronpedia

    INDEX
    Explanations

    code and data

    np_max-act · gemini-2.0-flash

    This neuron selectively activates on mentions of the “Search” tool—tokens like “Search,” “search engine,” “search query,” etc.—in the prompt’s tool description.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    Mission
    -0.07
     tablets
    -0.06
    /write
    -0.06
    Da
    -0.06
     preparedStatement
    -0.06
     humanities
    -0.06
     beb
    -0.06
     θέση
    -0.06
     пері
    -0.06
     infections
    -0.06
    POSITIVE LOGITS
     Northwest
    0.06
     فع
    0.06
     перш
    0.06
    DOCTYPE
    0.06
     Khal
    0.06
     whatsoever
    0.06
    ombo
    0.06
    (SQLException
    0.06
    ingo
    0.06
    <!
    0.06
    Act Density 0.024%

    No Known Activations