AI Paper Atlas

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

FiGURe: Simple and Efficient Unsupervised Node Representations with Filter Augmentations

Chanakya Ekbote, Ajinkya Deshpande, Arun Iyer, SUNDARARAJAN SELLAMANICKAM, Ramakrishna B Bairi

NeurIPS 2023 · Label-efficient and robust learning › Self-supervised and contrastive · approach: Empirical contrastive visual learning · 1 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 Random Fourier Feature projection at 32 dims (ablation)GRACE · chameleonaccuracy (%): 53.25 -> 67.39Improved
add RFF projection at 32 dims (ablation)DGI · chameleonaccuracy (%): 45.48 -> 56.01Improved
add RFF projection at 32 dims (ablation)FiGURe · squirrelaccuracy (%): 39.38 -> 48.89Improved
add RFF projection at 32 dims (ablation)FiGURe · coraaccuracy (%): 82.88 -> 82.56No clear effect
add RFF projection at 32 dims (ablation)SUGRL · coraaccuracy (%): 65.35 -> 70.06Improved
add RFF projection at 128 dims (ablation)MVGRL · coraaccuracy (%): 85.31 -> 84.61No clear effect
share encoder weights across filters instead of independent encoders (ablation)FiGURe · squirrelaccuracy (%): 50.52 -> 52.23Improved
share encoder weights across filters (ablation)FiGURe · citeseeraccuracy (%): 75.03 -> 74.77No clear effect
use 2-layer GCN encoder instead of 1 layer (ablation)FiGURe · chameleonaccuracy (%): 68.55 -> 53.53Hurt
use ANOVA kernel random features instead of RFF (ablation)FiGURe · coraaccuracy (%): 87.00 -> 83.26Hurt
use ChebNet K=11 filter bank instead of GPRGNN filters (ablation)FiGURe · coraaccuracy (%): 87.00 -> 76.14Hurt
use BernNet K=11 filters instead of K=3 (ablation)FiGURe · squirrelaccuracy (%): 37.07 -> 43.48Improved

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