# Preprint tests 4-bit quantization across hybrid LLM layers

_Published Friday, September 4, 2026 at 5:07 PM EDT · Science, AI · Latest · Tier 2 — Notable_

A preprint describes Minima, a 4-bit NVFP4 W4A4 quantization recipe applied to all 496 linear layers of a hybrid 27B language model, including its Gated DeltaNet layers. The authors report results near BF16 within seed noise across listed benchmarks, a 17.5 GiB model size, and 14% to 19% faster prefill versus compared recipes. They attribute the result to block scaling, gate behavior and recurrence noise dynamics.

## Sources

- [cs.AI updates on arXiv.org](https://arxiv.org/abs/2609.04098)

---
Canonical: https://techandbusiness.org/newswire/e1o8z_dP9byuEhn6Cibxwt
Published: 2026-09-04T21:07:05.215Z
Story chronology: 2026-09-04T04:00:00.000Z
Retrieved: 2026-10-20T04:30:42.095Z
Publisher: Tech & Business (techandbusiness.org)