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

    Neuronpedia

    INDEX
    Explanations

    principle followed by explanation

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     are
    0.98
    ру
    0.84
    ियों
    0.77
    пи
    0.77
    て
    0.76
    ми
    0.75
     is
    0.73
    ре
    0.73
    но
    0.71
    ての
    0.70
    POSITIVE LOGITS
    us
    0.86
    ad
    0.80
    ↵
    0.76
    น้อง
    0.75
    il
    0.74
    ir
    0.74
    H
    0.71
    Α
    0.71
    HAS
    0.70
    IS
    0.70
    Act Density 0.021%

    No Known Activations