AI Paper Atlas

Research map / Learning theory and algorithms / Combinatorial and search optimization

Design Linear Constrained Neural Layers with Implicit Convex Optimization

Junchi Yan, Jiaxi Liu, Yihui Tu, Fangyuan Zhou, Wenzheng Pan, Zhongteng Gui, Liangliang Shi

ICML 2026 · Learning theory and algorithms › Combinatorial and search optimization · approach: Constrained neural network optimization

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
replace Sinkhorn with BLCLayer-10 itersNGMv2 · Pascal VOC Keypoint partial graph matching (unfiltered)F1: 0.5453 -> 0.6012Improved
replace LinSAT-5 with BLCLayer-5 itersNGMv2 · Pascal VOC Keypoint partial graph matching (unfiltered)F1: 0.5858 -> 0.5972Improved
replace LinSAT-5 with BLCLayer-5 itersNGMv2 · Pascal VOC Keypoint partial graph matching (unfiltered)training time: 1h 56m -> 1h 28mImproved
raise temperature tau from 0.01 to 0.1 in LinSAT (ablation)NGMv2 · Pascal VOC Keypoint partial graph matching (unfiltered)F1: 0.5839 -> 0.5041Hurt
raise temperature tau from 0.01 to 0.1 in BLCNet (ablation)NGMv2 · Pascal VOC Keypoint partial graph matching (unfiltered)F1: 0.6012 -> 0.5921Hurt
increase LinSAT iterations from 10 to 40 (ablation)NGMv2 · Pascal VOC Keypoint partial graph matching (unfiltered)F1: 0.5839 -> 0.5894No clear effect
use GLCNet instead of LinSAT-100StemGNN · Portfolio allocation (S&P 500)Sharpe ratio: 2.755 -> 3.053Improved
use BLCNet-5 instead of LinSAT-100StemGNN · Portfolio allocation (S&P 500)training time: 4862.5 s -> 202.65 sImproved
add linear constraint layer vs unconstrainedStemGNN · Portfolio allocation (S&P 500)Sharpe ratio: 2.190 -> 3.053Improved
reduce LinSAT test temperature from 1/200 to 1/10 (ablation)LinSAT · Portfolio allocation (S&P 500), ablationSharpe ratio: 2.124 -> 1.311Hurt
increase GLCNet iterations from 5 to 10 (ablation)GLCNet · Portfolio allocation (S&P 500), ablationSharpe ratio: 3.000 -> 3.027No clear effect
use GLCNet instead of HardNetLP solver · Linear programming (3 mixed constraint scales)objective percentage: 0.908 -> 0.956 (500 vars); 0.869 -> 0.925 (2k vars)Improved

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