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

    Neuronpedia

    INDEX
    Explanations

    What

    np_max-act · gemini-2.0-flash

    The neuron detects words and phrases used when asking follow‐up questions or requesting more details (e.g. “what,” “please,” “provide,” “your,” “current,” “information,” “Additionally”).

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    	nil
    -0.07
    Kids
    -0.07
    FullScreen
    -0.07
     ประก
    -0.06
    still
    -0.06
     lagi
    -0.06
    iyoruz
    -0.06
    SOR
    -0.06
    итор
    -0.06
    ertime
    -0.06
    POSITIVE LOGITS
     "\"
    0.07
     '${
    0.06
    <const
    0.06
    BagConstraints
    0.06
     TextFormField
    0.06
    +++
    0.06
     ona
    0.06
     AttributeError
    0.06
     "${
    0.06
     engines
    0.06
    Act Density 0.040%

    No Known Activations