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

    Neuronpedia

    INDEX
    Explanations

    (

    np_max-act · gemini-2.0-flash

    This neuron fires on “special” tokens that don’t start with a lowercase letter—e.g. dotted method calls (.join, .extend), single‐letter capitals in acronyms (A, B, C, …), or accented/non-Latin initials.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8

    technical terms and concepts related to science and law.

    oai_token-act-pair · gpt-4o-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
     lovers
    -0.07
     empirical
    -0.07
     nombres
    -0.06
     Huang
    -0.06
    byterian
    -0.06
    वर
    -0.06
     Canvas
    -0.06
     Innovative
    -0.06
     nominations
    -0.06
     matters
    -0.06
    POSITIVE LOGITS
    _Call
    0.07
    平成
    0.07
     الخط
    0.07
     divul
    0.06
     Sé
    0.06
     Χα
    0.06
    _tipo
    0.06
     derecho
    0.06
    fun
    0.06
    /rec
    0.06
    Act Density 0.423%

    No Known Activations