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

    Neuronpedia

    INDEX
    Explanations

    even / not / never / barely

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     perfectly
    0.73
     perfect
    0.73
     exclusively
    0.68
    非常好
    0.68
     always
    0.67
     lots
    0.65
     automatically
    0.64
    เสมอ
    0.64
     exclusive
    0.64
     perfecta
    0.63
    POSITIVE LOGITS
     even
    1.73
     Even
    1.67
    Even
    1.65
    even
    1.61
     даже
    1.46
     حتی
    1.45
     überhaupt
    1.35
    哪怕
    1.33
     siquiera
    1.31
     Bahkan
    1.27
    Act Density 0.406%

    No Known Activations