Discussion

TurboQuant: Redefining AI efficiency with extreme compression

bluequbit: 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?

benob: 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: I think it is though- plenty of ‘it’s not X, it’s Y’ and AI seems to love corny section header names.

Maxious: https://mesuvash.github.io/blog/2026/turboquant-interactive/ has a little visualisation

mrugge: 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.

spencerflem: 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?

benob: Maybe they quantized a bit too much the model parameters...

moktonar: Aren’t polar coordinates still n-1 + 1 for radius for n-dim vector? If so I understand that angles can be quantized better but when radius r is big the error is large for highly quantized angles right? What am I missing?

amitport: r is a single value per vector. You don't have to quantize it, you can keep it and quantize the billion+ other coordinates of the vector.

vincnetas: i think grid can be a surface of the unit sphere