AI Paper Atlas

Research map / Learning theory and algorithms / Optimal transport

LaCoOT: Layer Collapse through Optimal Transport

Victor Quétu, Zhu Liao, Nour Hezbri, Fabio Pizzati, Enzo Tartaglione

ICCV 2025 · Learning theory and algorithms › Optimal transport · approach: Neural optimal transport · 2 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
raise regularization weight lambda from 0.01 to 5 (ablation)LaCoOT · ResNet-18 on CIFAR-10Top-1 accuracy after block removal: 73.04 -> 90.99Improved
add the Max-Sliced Wasserstein regularization (lambda>0) before ranking blocks (ablation)LaCoOT · ResNet-18 on CIFAR-10Top-1 vs MACsImproved
use LaCoOT importance score instead of block influence or random removal (ablation)LaCoOT · ResNet-18 on CIFAR-10Performance/compression trade-offImproved
use Sliced Wasserstein instead of Max-Sliced Wasserstein (ablation)LaCoOT (lambda=5) · ResNet-18 on CIFAR-10Top-1 vs MACsHurt
reduce number of projections to 1 (ablation)LaCoOT (lambda=5) · ResNet-18 on CIFAR-10Top-1 vs MACsHurt
reduce batch size used for the distance (ablation)LaCoOT (lambda=5) · ResNet-18 on CIFAR-10Top-1Hurt
replace MSWD with KL divergence (ablation)LaCoOT · ResNet-18 on CIFAR-10Top-1 vs MACsHurt
replace MSWD with l1, l2 or MMD (ablation)LaCoOT · ResNet-18 on CIFAR-10Top-1 vs MACsNo clear effect
add LaCoOT regularized finetuning to DiT-XL/2 with two blocks removed (ablation)DiT-XL/2 on ImageNet · ImageNet generationFID-50k: FID-50k twice as lowImproved
add DepGraph (0.2) on top of LaCoOT (ablation)LaCoOT (lambda=5) · ResNet-18 on CIFAR-10MACs / top-1: 64.69 -> 57.22 MACs, 90.99 -> 90.96 top-1Improved
add DepGraph (0.3) on top of LaCoOT (ablation)LaCoOT (lambda=5) · ResNet-18 on CIFAR-10Top-1: 90.99 -> 89.27Hurt
add regularization for 3x3 conv replaced by avgpool and 1x1 conv (lambda 0 to 0.1) (ablation)LaCoOT · ResNet-18 on CIFAR-10Performance loss vs full model: 0.98% -> 0.18% lossImproved

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