What surprised you about this? This seems pretty in line with optimization work seen from a wide range of models.

Did you try doing the same on Opus or others? How did they do?

I don't love these "look, this frontier model blew me away" comments when there is no point of comparison.

We tried it with Fable 5.1 as well, which was 3-4x slower than 5.2. We also have a blog with benchmark and details if you're curious https://discoveredmaterials.com/research/accelerating-quantu...