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

    Neuronpedia

    INDEX
    Explanations

    reaching zero or specific goals

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     almost
    0.61
     quase
    0.58
    Almost
    0.58
     Almost
    0.57
    almost
    0.55
     minus
    0.55
     prawie
    0.53
    minus
    0.50
     Minus
    0.50
     hampir
    0.49
    POSITIVE LOGITS
    깔
    0.41
     будете
    0.40
    Ꮭ
    0.40
    缄
    0.39
    िएगा
    0.39
    揮
    0.38
    closure
    0.38
     بمعنى
    0.38
     абсолю
    0.37
    犸
    0.37
    Act Density 0.000%

    No Known Activations