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

    Neuronpedia

    INDEX
    Explanations

    correctness, inaccuracy, and untruth

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
    autogui
    0.44
    carbam
    0.41
    ếc
    0.41
    MOS
    0.40
     раствор
    0.39
     सॉल्यूशन
    0.39
     ಸಮಸ್ಯ
    0.38
    tet
    0.38
     imaginable
    0.38
     હાથ
    0.36
    POSITIVE LOGITS
     inaccurate
    1.66
     accurate
    1.66
     untrue
    1.57
     inaccuracy
    1.55
    accurate
    1.49
     Accurate
    1.48
     inaccuracies
    1.47
     accuracy
    1.41
    准确
    1.30
     Accuracy
    1.29
    Act Density 0.027%

    No Known Activations