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

    Neuronpedia

    INDEX
    Explanations

    formulas, AI, rules

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    relevant
    0.40
    ወቅ
    0.39
     مناسب
    0.37
     ».
    0.37
    ಧ್ಯ
    0.36
    his
    0.36
    epochs
    0.36
     potrebné
    0.36
    当前
    0.35
    lovely
    0.35
    POSITIVE LOGITS
     enhances
    0.46
     WWE
    0.44
    简直
    0.43
    ✞
    0.42
    <0xF4>
    0.42
     EVERY
    0.42
     전통
    0.42
     throwback
    0.41
     doesn
    0.40
     are
    0.40
    Act Density 0.000%

    No Known Activations