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

    Neuronpedia

    INDEX
    Explanations

    Publication

    np_max-act · gemini-2.0-flash

    This neuron primarily detects references to patent publications (i.e. citations of patent document numbers).

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    years
    -0.07
     Love
    -0.06
    -face
    -0.06
     Take
    -0.06
    Chair
    -0.06
    _open
    -0.06
    .hardware
    -0.06
     SPECIAL
    -0.06
     Iraq
    -0.06
     Technology
    -0.06
    POSITIVE LOGITS
    иж
    0.07
    -double
    0.06
    구
    0.06
    岡
    0.06
    양
    0.06
     kinetic
    0.06
    っても
    0.06
    <Action
    0.06
     Bou
    0.06
    Cliente
    0.06
    Act Density 0.003%

    No Known Activations