AI Paper Atlas

Research map / Label-efficient and robust learning / Self-supervised and contrastive

Can local learning match self-supervised backpropagation?

Wu S. Zihan, Ariane Delrocq, Wulfram Gerstner, Guillaume Bellec

ICML 2026 · Label-efficient and robust learning › Self-supervised and contrastive · approach: Data-efficient and robust self-supervision

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 2D spatial dependence to feedback B (ablation)CLAPP++ · CIFAR-10 linear probeaccuracy: 73.21 -> 80.51Improved
add 2D spatial dependence to feedback B (ablation)CLAPP++ · STL-10 linear probeaccuracy: 75.10 -> 78.66Improved
add 2D spatial dependence to feedback B (ablation)CLAPP++ · Tiny-ImageNet linear probeaccuracy: 28.18 -> 36.63Improved
add 2D spatial dependence to feedback B (ablation)CLAPP++ · ImageNet linear probeaccuracy: 38.31 -> 42.55Improved
add direct feedback from last layer (ablation)CLAPP++ · STL-10 linear probeaccuracy: 78.66 -> 79.38Improved
add direct feedback from last layer (ablation)CLAPP++ · ImageNet linear probeaccuracy: 42.55 -> 44.16Improved
optimise sum of direct-feedback and local losses (both) (ablation)CLAPP++ · CIFAR-10 linear probeaccuracy: 80.51 -> 81.18Improved
optimise sum of direct-feedback and local losses (both) (ablation)CLAPP++ · ImageNet linear probeaccuracy: 42.55 -> 42.49No clear effect
use local-SSL instead of end-to-end BP with same lossCLAPP++ · ImageNet linear probeaccuracy: 48.52 -> 42.55Hurt
use local-SSL instead of BPCLAPP++ · VRAM, STL-10 (6-layer VGG)peak VRAM (GB): 5.47 -> 3.38Improved
use local-SSL instead of BPCLAPP++ · VRAM, ImageNet (8-layer VGG)peak VRAM (GB): 11.10 -> 4.60Improved
add spatial dependence to BP-SSL (ablation)BP-CLAPP++ · benchmarksaccuracyNo 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