ICLR 2023 accepted papers on Graph learning
110 of the 1,575 papers accepted at ICLR 2023 (7.0%) are on Graph learning, compared with 7.4% at ICLR 2022. They fall into 4 topics.
See ICLR 2023 in the interactive map
Working on something in this area? Describe your idea in scime atlas to see the closest accepted papers by meaning and how crowded the spot has become.
Topics and papers
GNN architectures · 55 papers
- Equiformer: Equivariant Graph Attention Transformer for 3D Atomistic Graphs (67 citations)
- Sign and Basis Invariant Networks for Spectral Graph Representation Learning (24 citations)
- A critical look at the evaluation of GNNs under heterophily: Are we really making progress? (21 citations)
- NAGphormer: A Tokenized Graph Transformer for Node Classification in Large Graphs (21 citations)
- Ordered GNN: Ordering Message Passing to Deal with Heterophily and Over-smoothing (18 citations)
- Graph Neural Networks for Link Prediction with Subgraph Sketching (17 citations)
- Graph Neural Networks are Inherently Good Generalizers: Insights by Bridging GNNs and MLPs (16 citations)
- Specformer: Spectral Graph Neural Networks Meet Transformers (16 citations)
- Rethinking the Expressive Power of GNNs via Graph Biconnectivity (15 citations)
- Anti-Symmetric DGN: a stable architecture for Deep Graph Networks (13 citations)
- CktGNN: Circuit Graph Neural Network for Electronic Design Automation (13 citations)
- MLPInit: Embarrassingly Simple GNN Training Acceleration with MLP Initialization (12 citations)
and 43 more in the interactive map.
Graph self-supervised learning · 22 papers
- Mole-BERT: Rethinking Pre-training Graph Neural Networks for Molecules (135 citations)
- LightGCL: Simple Yet Effective Graph Contrastive Learning for Recommendation (99 citations)
- DiGress: Discrete Denoising diffusion for graph generation (73 citations)
- DIFFormer: Scalable (Graph) Transformers Induced by Energy Constrained Diffusion (26 citations)
- Learning on Large-scale Text-attributed Graphs via Variational Inference (25 citations)
- Spectral Augmentation for Self-Supervised Learning on Graphs (16 citations)
- Graph Contrastive Learning for Skeleton-based Action Recognition (13 citations)
- Automated Data Augmentations for Graph Classification (9 citations)
- Energy-based Out-of-Distribution Detection for Graph Neural Networks (8 citations)
- Edgeformers: Graph-Empowered Transformers for Representation Learning on Textual-Edge Networks (5 citations)
- Link Prediction with Non-Contrastive Learning (5 citations)
- Confidence-Based Feature Imputation for Graphs with Partially Known Features (3 citations)
and 10 more in the interactive map.
Graph algorithms and clustering · 17 papers
- Learning to Induce Causal Structure (17 citations)
- On Representing Linear Programs by Graph Neural Networks (9 citations)
- On Representing Mixed-Integer Linear Programs by Graph Neural Networks (6 citations)
- Fuzzy Alignments in Directed Acyclic Graph for Non-Autoregressive Machine Translation (5 citations)
- Random Laplacian Features for Learning with Hyperbolic Space (1 citation)
- Robust Scheduling with GFlowNets (1 citation)
- AutoTransfer: AutoML with Knowledge Transfer - An Application to Graph Neural Networks
- DAG Learning on the Permutahedron
- Graph Signal Sampling for Inductive One-Bit Matrix Completion: a Closed-form Solution
- Unveiling the sampling density in non-uniform geometric graphs
- Graph-based Deterministic Policy Gradient for Repetitive Combinatorial Optimization Problems
- Neural DAG Scheduling via One-Shot Priority Sampling
and 5 more in the interactive map.
Knowledge and temporal graphs · 16 papers
- Do We Really Need Complicated Model Architectures For Temporal Networks? (19 citations)
- Neural Compositional Rule Learning for Knowledge Graph Reasoning (15 citations)
- ExpressivE: A Spatio-Functional Embedding For Knowledge Graph Completion (12 citations)
- TILP: Differentiable Learning of Temporal Logical Rules on Knowledge Graphs (11 citations)
- Hierarchical Relational Learning for Few-Shot Knowledge Graph Completion (9 citations)
- Relational Attention: Generalizing Transformers for Graph-Structured Tasks (6 citations)
- EAGLE: Large-scale Learning of Turbulent Fluid Dynamics with Mesh Transformers (5 citations)
- Logical Message Passing Networks with One-hop Inference on Atomic Formulas (4 citations)
- Towards Open Temporal Graph Neural Networks (4 citations)
- Logical Entity Representation in Knowledge-Graphs for Differentiable Rule Learning (3 citations)
- Learning Heterogeneous Interaction Strengths by Trajectory Prediction with Graph Neural Network (2 citations)
- Direct Embedding of Temporal Network Edges via Time-Decayed Line Graphs (1 citation)
and 4 more in the interactive map.