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

    Neuronpedia

    INDEX
    Explanations

    in or for

    np_max-act · gemini-2.0-flash

    The neuron is sensitive to tokens used for directly addressing the reader or interlocutor, especially second-person pronouns (e.g. “you,” “du,” “Ihr”) and related formality markers.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
     Blacks
    -0.07
    白
    -0.06
    CLK
    -0.06
     Apps
    -0.06
     στους
    -0.06
     빨
    -0.06
     面
    -0.06
     igen
    -0.06
     Stevens
    -0.06
     Townsend
    -0.06
    POSITIVE LOGITS
    .Sqrt
    0.08
     
    0.07
    ="(
    0.06
    _math
    0.06
    lei
    0.06
     усп
    0.06
     renew
    0.06
    LEGAL
    0.06
    .den
    0.06
    =P
    0.06
    Act Density 0.072%

    No Known Activations