TurboQuant: Redefining AI efficiency with extreme compression

(research.google)

91 points | by ray__ 2 hours ago

3 comments

  • benob 28 minutes ago
    This is the worst lay-people explanation of an AI component I have seen in a long time. It doesn't even seem AI generated.
    • spencerflem 27 minutes ago
      I think it is though-

      “ TurboQuant, QJL, and PolarQuant are more than just practical engineering solutions; they’re fundamental algorithmic contributions backed by strong theoretical proofs. These methods don't just work well in real-world applications; they are provably efficient and operate near theoretical lower bounds.”

      • benob 22 minutes ago
        Maybe they quantized a bit too much the model parameters...
  • bluequbit 49 minutes ago
    I did not understand what polarQuant is.

    Is is something like pattern based compression where the algorithm finds repeating patterns and creates an index of those common symbols or numbers?

    • mrugge 34 minutes ago
      1. Efficient recursive transform of kv embeddings into polar coordinates 2. Quantize resulting angles without the need for explicit normalization. This saves memory via key insight: angles follow a distribution and have analytical form.
      • quotemstr 16 minutes ago
        Reminds me vaguely of Burrows-Wheeler transformations in bzip2.
    • Maxious 32 minutes ago
      • spencerflem 20 minutes ago
        I like the visualization, but I don’t understand the grid quantization. If every point is on the unit circle aren’t all the center grid cords unused?
  • hikaru_ai 23 minutes ago
    [dead]