AI Paper Atlas

Research map / Graph learning / Graph self-supervised learning

Structure-Aware Graph Hypernetworks for Neural Program Synthesis

Wenhao Li, Yudong Xu, Elias Boutros Khalil, Scott Sanner

ICLR 2026 · Graph learning › Graph self-supervised learning · approach: Graph domain adaptation

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 structure-aware message passing (Meta-GNN vs Meta-MLP)hypernetwork · SUMFIRST-n OOD testaccuracy: 0.145 -> 0.866Improved
add group-aware parameter tying (Meta-gMLP vs Meta-MLP)hypernetwork · SUMFIRST-n OOD testaccuracy: 0.145 -> 0.344Improved
synthesize weights with meta-learner instead of direct predictionMLP · SUMFIRST-n OOD testaccuracy: 0.140 -> 0.866Improved
synthesize weights with Meta-MLP instead of direct TransformerTransformer · ADDMOD-p OOD testaccuracy: 0.682 -> 0.947Improved
use Meta-GNN instead of Meta-MLPhypernetwork · ADDMOD-p OOD testaccuracy: 0.947 -> 0.943No clear effect
use Meta-GNN instead of direct TransformerTransformer · 1D-ARC mixture OOD testaccuracy: 0.603 -> 0.713Improved
use Meta-GNN instead of Meta-MLPhypernetwork · 1D-ARC Flip OODaccuracy: 0.127 -> 0.898Improved
use meta-learners instead of direct TransformerTransformer · 1D-ARC MoveRight OODaccuracy: 0.000 -> 0.996Improved
use meta-learners instead of direct TransformerTransformer · 1D-ARC RecolorBar OODaccuracy: 0.016 -> 0.000Hurt
use attention aggregation instead of average (ablation)Meta-GNN · ADDMOD-p small ablationaccuracy: 0.949 -> 0.976Improved
use attention aggregation instead of sum (ablation)Meta-GNN · ADDMOD-p small ablationaccuracy: 0.947 -> 0.976Improved
use separate parameters for reverse edges instead of shared (ablation)Meta-GNN · ADDMOD-p small ablationaccuracy: 0.976 -> 0.979Improved

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