Superseded baseline#64 of 80 most-superseded
SFMP
SFMP: Fine-Grained, Hardware-Friendly and Search-Free Mixed-Precision Quantization for Large Language ModelsLLM quantization · first seen Feb 1, 2026
superseded — cited as a baseline and beaten by newer methods
0 papers critique it · 1 beat it on benchmarks
Beaten on benchmarks
Head-to-head results where a newer method reports beating SFMP. Values are copied from the source paper's tables — verify against the cited paper.
- SignRoundV2: Closing the Performance Gap in Extremely Low-Bit Post-Training Quantization for LLMs
SignRoundV2 beats SFMP · Avg [Llama3.1-8B, Avg Bits 2.5]
68.57 vs 64.34
- SignRoundV2: Closing the Performance Gap in Extremely Low-Bit Post-Training Quantization for LLMs
SignRoundV2 beats SFMP · Avg [Qwen3-8B, Avg Bits 2.5]
66.82 vs 66.16
Newer alternatives
Recent methods in the same sub-problem, not yet superseded in the knowledge base.