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

    Neuronpedia

    INDEX
    Explanations

    technical classifications/descriptions

    np_max-act · gemini-2.0-flash

    This neuron activates on section‐heading phrases that introduce classification criteria—especially headings beginning with “By …” (e.g. “By affected component”).

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    Revision
    -0.07
     prognosis
    -0.07
     bottle
    -0.07
     значения
    -0.07
    History
    -0.06
     Rings
    -0.06
    Parking
    -0.06
     [--
    -0.06
     sociedad
    -0.06
     libertine
    -0.06
    POSITIVE LOGITS
     beim
    0.08
     فق
    0.07
    _VERIFY
    0.07
     clears
    0.06
     taşın
    0.06
     답
    0.06
    .poly
    0.06
    ionale
    0.06
     kolem
    0.06
    ственное
    0.06
    Act Density 0.150%

    No Known Activations