Why Gated DeltaNet Survives 4-Bit Quantization: NVFP4 W4A4 for the Recurrent Half of a Hybrid 27B LLM Paper • 2609.04098 • Published 1 day ago • 49
Why Gated DeltaNet Survives 4-Bit Quantization: NVFP4 W4A4 for the Recurrent Half of a Hybrid 27B LLM Paper • 2609.04098 • Published 1 day ago • 49
Minima-KV: Retention-Preserving KV Cache Compression with Mixed-Format Paged Attention Paper • 2608.23834 • Published 11 days ago
A Practical Tensor-Network Compression Pipeline for Production-Scale Large Language Models Paper • 2602.01613 • Published Feb 2