AI Paper Atlas

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

Emergent Visual Representations through Unsupervised Spiking Networks with Synaptic Pruning

Di Hong, Dazhong Rong, Yueming Wang

ICML 2026 · Label-efficient and robust learning › Self-supervised and contrastive · approach: Self-supervised neuroimaging representations

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 synaptic pruning (ablation)UL SNN · Allen Brain Mouse (CIFAR-10, ResNet18)RSA: 0.4279 -> 0.4608Improved
add synaptic pruning (ablation)UL SNN · Allen Brain Mouse (CIFAR-10, ResNet18)SVCCA: 0.6294 -> 0.6308Improved
add synaptic pruning (ablation)UL SNN · Macaque Face (CIFAR-10, ResNet18)SVCCA: 0.4576 -> 0.4687Improved
add synaptic pruning (ablation)UL SNN · Macaque Synthetic (CIFAR-10, ResNet34)TSVDR: 0.3284 -> 0.3619Improved
add synaptic pruning (ablation)UL SNN · Macaque Synthetic (Tiny-ImageNet, ResNet18)TSVDR: 0.3556 -> 0.3668 (below SL ANN 0.4048)Mixed
add synaptic pruning (ablation)UL SNN · Macaque Synthetic (CIFAR-10, ResNet18)SVCCA: 0.3194 -> 0.3226 (SL SNN 0.3461)Mixed
use unsupervised contrastive instead of supervised training (SNN) (ablation)SNN · Allen Brain Mouse (CIFAR-10, ResNet18)RSA: 0.1601 -> 0.4279Improved
use spiking network instead of ANN (unsupervised) (ablation)UL · Allen Brain Mouse (CIFAR-10, ResNet18)RSA: 0.1849 -> 0.4279Improved
use unsupervised instead of supervised (ANN) (ablation)ANN · Allen Brain Mouse (CIFAR-10, ResNet18)SVCCA: 0.6075 -> 0.6046No clear effect
add synaptic pruning (ablation)UL SNN · Few-shot CIFAR-10 transfer (20% labels)Top-1 accuracy: 62.94 -> 63.11Improved
add synaptic pruning (ablation)UL SNN · Gaussian noise 20%Top-1 accuracy: 59.07 -> 59.94Improved
add synaptic pruning (ablation)UL SNN · Gaussian noise 80%Top-1 accuracy: 52.44 -> 52.57No 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