AI Paper Atlas

Research map / Architectures and efficiency / Attention and transformers

Breaking the Correlation Plateau: On the Optimization and Capacity Limits of Attention-Based Regressors

Jingquan Yan, Yuwei Miao, Peiran Yu, Junzhou Huang

ICLR 2026 · Architectures and efficiency › Attention and transformers · approach: Sparse and structured attention · 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
add ECA to ALMT (with dispersion-normalized PCC loss)ALMT · MOSIPCC: 0.783 -> 0.806Improved
add a plain PCC loss to ALMTALMT · MOSIF1: 0.851 -> 0.834Hurt
add ECA to EGNEGN · 10xProteomic (3-fold)PCC@M: 0.629 -> 0.716Improved
add ECA to FT-TransformerFT-Transformer · Appliance (UCI)PCC: 0.556 -> 0.598Improved
add ECA to FT-TransformerFT-Transformer · Online News (UCI)PCC: 0.408 -> 0.420Improved
add ECA to FT-TransformerFT-Transformer · Superconductivity (UCI)PCC: 0.920 -> 0.930Improved
add DATS (dispersion-aware temperature softmax) (ablation)ECA · Appliance (UCI)PCC: 0.561 -> 0.598Improved
add SRA (scaled residual aggregation) (ablation)ECA · Online News (UCI)PCC: 0.410 -> 0.420Improved
add DNPL (dispersion-normalized PCC loss) (ablation)ECA · Superconductivity (UCI)PCC: 0.922 -> 0.930Improved
add SRA to ECA (ablation)ECA · Appliance (UCI)MSE (x10^3): 5.994 -> 5.790Improved
add DATS to ECA (ablation)ECA · Online News (UCI)PCC: 0.418 -> 0.420No clear effect

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