arXiv cs.AIPaper
Why Gated DeltaNet Survives 4-Bit Quantization: NVFP4 W4A4 for the Recurrent Half of a Hybrid 27B LLM
The practical ceiling on quantization for hybrid architectures just moved higher. If you're deploying Qwen3.8-27B or similar hybrids, this says you can push to 4-bit across the entire stack and still match BF16 baseline. The mechanism study—why block scaling solves recurrent accumulation—is engineering guidance you can apply to your own quantization pipeline.