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

    Neuronpedia

    INDEX
    Explanations

    most or least ever

    np_acts-logits-general · gemini-2.5-flash-lite
    New Auto-Interp
    Negative Logits
     conviene
    0.38
    dyž
    0.37
    yı
    0.35
    uminação
    0.35
    сибо
    0.33
    hr
    0.33
    nabla
    0.33
     лучший
    0.32
    }")
    0.32
     prieš
    0.31
    POSITIVE LOGITS
     imaginable
    0.69
     ever
    0.65
     EVER
    0.62
     überhaupt
    0.54
    Ever
    0.49
    之一
    0.48
    ever
    0.45
     Ever
    0.45
    EVER
    0.44
     available
    0.41
    Act Density 0.051%

    No Known Activations