Unverified60% confidenceFactExact time
IQ4_XS is a non-uniform 4-bit quantization format in the llama.cpp ecosystem that uses a more fine-grained scaling factor allocation strategy within quantization blocks.
2
Sources
60%
Confidence
Long-term
Relevance
9/18/2026
First Seen
Sources
Related Entities
Related Claims
UnverifiedThe IQ series is a quantization scheme in llama.cpp targeting extremely low bit-rate (2–3 bit) scenarios, introducing non-uniform quantization grids based on importance matrices, vector quantization, and finer super-block structures79% similarUnverifiedQ4_K_M是llama.cpp项目引入的一种混合量化策略，对更敏感的层保留较高精度，对其余层使用更激进的压缩68% similarUnverifiedMXFP4 belongs to the MX (Microscaling) series of standards and is a block floating-point quantization specification in which a group of weights shares a single scaling factor.65% similarUnverifiedNF4 (4-bit NormalFloat) is an information-theoretically optimal data type proposed in the QLoRA paper (Dettmers et al., 2023).64% similarVerifiedINT4 quantization can theoretically compress a model's VRAM footprint to one-quarter of FP16, with commonly used tools including GPTQ, AWQ, and GGUF.63% similar
Cite This Claim
Stable URI
https://kongchang.com/claim/932906API
curl https://kongchang.com/api/v1/knowledge/claims/932906MCP
get_claim(id=932906)