AI Paper Atlas

Research map / Architectures and efficiency / Quantization

AdaSVD: Singular Value Decomposition with Adaptive Mechanisms for Large Multimodal Models

Zhiteng Li, Mingyuan Xia, Jingyuan Zhang, Zheng Hui, Haotong Qin, Linghe Kong, Yulun Zhang, Xiaokang Yang

CVPR 2026 · Architectures and efficiency › Quantization · approach: Post-training quantization of ViTs

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 adaComp (adaptive compensation) (ablation)AdaSVD on LLaMA2-7B · WikiText-2, 60% compressionperplexity: 78.82 -> 50.33Improved
add adaComp (adaptive compensation) (ablation)AdaSVD on LLaMA2-7B · WikiText-2, 50% compressionperplexity: 30.00 -> 25.58Improved
add adaCR (adaptive compression ratio vs constant) (ablation)AdaSVD on LLaMA2-7B · WikiText-2, 60% compressionperplexity: 69.46 -> 50.33Improved
add adaCR (adaptive compression ratio vs constant) (ablation)AdaSVD on LLaMA2-7B · WikiText-2, 50% compressionperplexity: 27.33 -> 25.58Improved
increase adaComp iterations from 1 to 15 (ablation)AdaSVD on LLaMA2-7B · WikiText-2, 60% compressionperplexity: 50.33 -> 62.34Hurt
increase adaComp iterations from 1 to 15 (ablation)AdaSVD on LLaMA2-7B · WikiText-2, 40% compressionperplexity: 14.76 -> 15.84Hurt
raise minimum retention ratio from 0.30 to 0.40 (ablation)AdaSVD on LLaMA2-7B · WikiText-2, 60% compressionperplexity: 50.33 -> 60.08Hurt
raise minimum retention ratio from 0.40 to 0.50 (ablation)AdaSVD on LLaMA2-7B · WikiText-2, 40% compressionperplexity: 15.01 -> 14.76No clear effect
use Moore-Penrose pseudoinverse update and stack-of-batch calibration (ablation)AdaSVD on LLaMA2-7B · compression errorupdate errorImproved
combine with GPTQ-INT4 quantization (ablation)AdaSVD on LLaMA2-7B · WikiText-2, 40% compressionperplexity: 14.76 -> 22.55Hurt
use AdaSVD instead of SVD-LLM, combined with GPTQ-INT4AdaSVD on LLaMA2-7B · WikiText-2, 40% compressionperplexity: 33.56 -> 22.55Improved
use AdaSVD instead of SVD-LLMAdaSVD on LLaMA2-7B · WikiText-2, 60% compressionperplexity: 89.90 -> 50.33Improved

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