[stingraycharles]

TurboQuant went through ICLR review, has multiple Google Research co-authors, open-source implementations, CUDA kernels, and LongBench benchmarks.

Contrast that with your paper: no experiments, no implementation, no empirical validation of any kind.

Did you try engaging with LLM researchers and get their feedback on your paper?

[mskkm]

went through ICLR review: scores 4 4 6 10, serious? open-source implementations: where is the official code? CUDA kernels: where?

[EGreg]

Since yesterday, I put up the source code btw:

https://github.com/Safebots/KV