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

    Neuronpedia

    INDEX
    Explanations

    Japanese, Thai, Korean sentence endings

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    具有
    0.63
    prim
    0.62
    逐
    0.54
    ents
    0.54
    最も
    0.53
    使其
    0.53
     simult
    0.52
    Amid
    0.52
    異なる
    0.52
    将
    0.52
    POSITIVE LOGITS
    です
    1.46
    ですね
    1.44
    嗎
    1.43
    吗
    1.41
    ですか
    1.41
    啊
    1.39
    นะ
    1.37
    입니다
    1.37
    ค่ะ
    1.35
    이에요
    1.34
    Act Density 0.026%

    No Known Activations