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

    Neuronpedia

    INDEX
    Explanations

    ably

    np_max-act · gemini-2.0-flash

    This neuron detects deontic or instructional modal language conveying obligations and rules (e.g. “must,” “should,” “never,” “always”).

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
    select
    -0.07
     �
    -0.06
    -0.06
    ‌اش
    -0.06
    TimeZone
    -0.06
    Rs
    -0.06
     ==
    -0.06
    .mar
    -0.06
    Creates
    -0.06
    另
    -0.06
    POSITIVE LOGITS
     rally
    0.07
     영화
    0.07
    unity
    0.06
     massa
    0.06
    /per
    0.06
     facilitating
    0.06
    -purpose
    0.06
     purpose
    0.06
     моч
    0.06
    otive
    0.06
    Act Density 0.041%

    No Known Activations