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?