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

    Neuronpedia

    INDEX
    Explanations

    it

    np_max-act · gemini-2.0-flash

    The neuron preferentially activates on personal-perspective words and phrases (e.g. I/you/it, time to start, new one, New Jersey), effectively flagging first-person narration or direct address.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
     Cylinder
    -0.07
     Mikhail
    -0.06
    [param
    -0.06
    overview
    -0.06
     ประก
    -0.06
    ेव
    -0.06
    PullParser
    -0.06
     maxY
    -0.06
    (off
    -0.06
    _THRESHOLD
    -0.06
    POSITIVE LOGITS
     nasty
    0.07
     nast
    0.07
    .setData
    0.06
     DIN
    0.06
    -center
    0.06
    //------------------------------------------------------------------------------↵
    0.06
    :left
    0.06
     fp
    0.06
     recording
    0.06
     governed
    0.06
    Act Density 0.240%

    No Known Activations