AI Paper Atlas

Research map / Architectures and efficiency / Pruning and efficient tuning

A3: an Analytical Low-Rank Approximation Framework for Attention

Jeffrey T. H. Wong, Cheng Zhang, Xinye Cao, Pedro Gimenes, Christos-Savvas Bouganis, George Anthony Constantinides, Wayne Luk, Yiren Zhao

ICML 2026 · Architectures and efficiency › Pruning and efficient tuning · approach: Low-rank compression and merging · 0 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
use A3 instead of SVD-LLMA3 · WikiText-2 perplexity, LLaMA-3.1-70B (10% compression)perplexity: 7.87 -> 4.69Improved
use A3 instead of SVD-LLMA3 · WikiText-2 perplexity, LLaMA-3.1-8B (20% compression)perplexity: 42.28 -> 11.36Improved
use A3 instead of SVD-LLMA3 · Downstream average accuracy, LLaMA-3.1-70B (10% compression)accuracy: 0.6797 -> 0.7508Improved
use A3 instead of SVD-LLM at higher compressionA3 · C4 perplexity, LLaMA-3.1-70B (20% compression)perplexity: 13.77 -> 13.94Hurt
use A3 instead of SVD-LLM at 20% compressionA3 · Downstream accuracy, LLaMA-3.1-70B (20% compression, ARC-c)accuracy: 0.4957 -> 0.4667Hurt
use A3 instead of Clover (MPT-7B, 20%)A3 · WikiText-2 perplexity, MPT-7Bperplexity: 77.78 -> 8.05Improved
use A3 instead of Palu (MPT-7B, 60%)A3 · WikiText-2 perplexity, MPT-7B (60% compression)perplexity: 25.09 -> 15.58Improved
use A3 instead of Palu (MPT-7B, 80%)A3 · WikiText-2 perplexity, MPT-7B (80% compression)perplexity: 13714 -> 849Improved
use A3 instead of SVD-LLM with SlimPajama calibration (LLaMA-7B, 40% compression)A3 · Average WikiText-2/C4 perplexity, LLaMA-7B (40%)perplexity: 30.84 -> 25.22Improved
use A3 instead of SVD-LLM with WikiText-2 calibration (LLaMA-7B, 40%)A3 · Average WikiText-2/C4 perplexity, LLaMA-7B (40%)perplexity: 44.58 -> 12.82Improved
use activation autocorrelation weighting (A3-QK) instead of identity (CLOVER) (ablation)A3-QK · MPT-7B QK component (20% compression)delta PPLImproved
use A3-OV instead of Palu (ablation)A3-OV · MPT-7B OV component (20% compression)delta PPLImproved

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