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

    Neuronpedia

    INDEX
    Explanations

    Korean and Japanese

    np_max-act · gemini-2.0-flash

    The neuron is primarily activated by Korean grammatical case or topic particles (especially the one‐character particles like "이" and "은") that follow nouns.

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
     خدم
    -0.07
    ям
    -0.07
    fly
    -0.07
    -disc
    -0.06
    _PROVID
    -0.06
    -control
    -0.06
     broadcasting
    -0.06
     суп
    -0.06
    cart
    -0.06
    .Cookie
    -0.06
    POSITIVE LOGITS
     이
    0.09
    ByName
    0.07
    本
    0.07
     그
    0.07
     本
    0.07
     この
    0.07
    、その
    0.07
    거
    0.06
     이번
    0.06
    이
    0.06
    Act Density 0.012%

    No Known Activations