AI Paper Atlas

Research map / Architectures and efficiency / Quantization

FLRQ: Faster LLM Quantization with Flexible Low-Rank Matrix Sketching

Hongyaoxing Gu, Lijuan Hu, Shuzi Niu, Fangfang Liu

AAAI 2026 · Architectures and efficiency › Quantization · approach: Post-training quantization of LLMs · 2 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add BLC iterative low-rank/clipping refinement (ablation)FLRQ · WikiText2 W2A16perplexity: 2.1e6 -> 9.14 (LLaMA2-7b)Improved
add BLC (ablation)FLRQ · WikiText2 W4A16perplexity: 5.55 -> 5.55 (LLaMA2-7b)No clear effect
use flexible rank (R1-FLR) instead of fixed rank 64 (ablation)FLRQ · WikiText2 W4A16 LLaMA2-7bperplexity: 5.73 -> 5.74No clear effect
use FLRQ vs OmniQuant at 2-bitLLaMA2-7b · WikiText2 W2A16perplexity: 11.06 -> 9.14Improved
use FLRQ vs OmniQuant at 4-bitLLaMA2-7b · WikiText2 W4A16perplexity: 5.58 -> 5.55No clear effect
FLRQ with flexible avg rank ~39 vs LQER rank 256LLaMA2-7b · WikiText2 W2A16perplexity: 10.33 -> 9.14Improved
FLRQ vs LQER on C4 at 2-bitLLaMA2-7b · C4 W2A16perplexity: 12.12 -> 12.10Mixed
use R1-Sketch instead of exact SVD (ablation)FLRQ · WikiText2 W3A16perplexity: 15.53 -> 15.53 (OPT-1.3B); time 8.7m -> 6.1mNo clear effect
increase sketch iterations beyond 2 (ablation)FLRQ · WikiText2 W3A16perplexity: 15.53 -> 15.53 (it 2 -> 8)No clear effect
add power iterations (it=0 to 2) (ablation)R1-Sketch · WikiText2 W3A16 OPT-1.3Bperplexity: 16.84 -> 15.53Improved
use FLRQ vs OmniQuant quantization time at 2-bitLLaMA2-7b · quantization timehours: 3.1h -> 2.0hImproved
use FLRQ vs OmniQuant zero-shot at 2-bitLLaMA2-7b · 6-task zero-shot avgaccuracy: 50.1% -> 56.4%Improved

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 2 of 2 on most shared results.

Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy