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

    Neuronpedia

    INDEX
    Explanations

    division by 2 or 3

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     одном
    0.49
    ionic
    0.46
     одну
    0.45
    medium
    0.42
    一つ
    0.39
     one
    0.38
    safe
    0.37
    mill
    0.37
     één
    0.37
    க்கொண்ட
    0.37
    POSITIVE LOGITS
     Suites
    0.42
     fours
    0.41
     fives
    0.40
     Divided
    0.40
    দ্যাল
    0.39
    IDES
    0.39
     drei
    0.37
     собой
    0.37
     four
    0.37
     threes
    0.37
    Act Density 0.051%

    No Known Activations