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

    Neuronpedia

    INDEX
    Explanations

    Korean sentence endings

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     choked
    0.89
    נים
    0.85
    ）。
    0.85
    енты
    0.85
     چه
    0.83
    inspiring
    0.83
     clouded
    0.82
     gripped
    0.80
    」。
    0.79
    𝓖
    0.79
    POSITIVE LOGITS
     만
    1.16
    만
    1.12
     사용
    1.09
    .
    1.09
     중
    1.08
     주
    1.03
     적
    1.03
     또한
    1.03
     위치
    1.02
    에
    1.01
    Act Density 0.002%

    No Known Activations