AI Paper Atlas

Research map / Architectures and efficiency / Quantization

First-Order Error Matters: Accurate Compensation for Quantized Large Language Models

Xingyu Zheng, Haotong Qin, Yuye Li, Haoran Chu, Jiakai Wang, Jinyang Guo, Michele Magno, Xianglong Liu

AAAI 2026 · Architectures and efficiency › Quantization · approach: Post-training quantization of LLMs · 6 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 first-order error term to GPTQ compensationGPTQ · Llama3-8B WikiText2 W3perplexity: 9.86 -> 8.32Improved
FOEM instead of GPTAQGPTAQ · Llama3-8B MMLU 5-shot W3accuracy: 53.8 -> 56.1Improved
add first-order term to GPTQGPTQ · Llama2-7B MMLU W3accuracy: 40.3 -> 42.0Improved
add first-order term at 4-bitGPTQ · Llama3-8B WikiText2 W4perplexity: 6.54 -> 6.54No clear effect
add first-order term at 4-bitGPTAQ · Llama3.2-3B WikiText2 W4perplexity: 8.66 -> 8.67No clear effect
FOEM vs GPTQ on Llama3-70BGPTQ · Llama3-70B WikiText2 W3perplexity: 5.37 -> 5.36No clear effect
FOEM vs GPTQ on 3-bit zero-shot BoolQGPTQ · Llama3-8B BoolQ W3accuracy: 71.8 -> 69.1Hurt
combine FOEM with SpinQuant at W4A4KV4GPTQ · Llama3-8B WikiText2perplexity: 8.55 -> 8.35Improved
FOEM on Qwen3-8BGPTQ · Qwen3-8B WikiText2 W3perplexity: 8.32 -> 8.17Improved
FOEM on Mamba-1.4BGPTAQ · Mamba-1.4B WikiText2 W3A16perplexity: 14.10 -> 13.91Improved
raise stabilization factor beta from 0.1 to 1.0 (ablation)FOEM · Llama3-8B WikiText2 W3perplexity: 8.32 -> 10.37Hurt
raise beta from 0.1 to 0.3 (ablation)FOEM · Llama3-8B 0-shot avg W3accuracy: 65.5 -> 65.8No clear effect

Earlier methods this paper compares with

Named baselines from earlier accepted papers in this paper's results tables: better than 1 of 1 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