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

    Neuronpedia

    INDEX
    Explanations

    ことが

    np_max-act · gemini-2.0-flash

    This neuron detects the Japanese polite potential construction “～ことができます” (expressions of “can/able to”).

    oai_token-act-pair · o4-miniTriggered by @xinyanhu8
    New Auto-Interp
    Negative Logits
     Rug
    -0.07
     smarter
    -0.06
    beta
    -0.06
     Ajax
    -0.06
    jc
    -0.06
     Sour
    -0.06
     mua
    -0.06
    "f
    -0.06
     Submission
    -0.06
     Sne
    -0.06
    POSITIVE LOGITS
    ことが
    0.12
    ことは
    0.09
    ことも
    0.09
    こと
    0.07
    的に
    0.07
    ofilm
    0.07
    クロ
    0.07
     것으로
    0.07
    krát
    0.07
     Bağ
    0.07
    Act Density 0.004%

    No Known Activations