| Graph Neural Reaction Diffusion Models | | 94.59 ± 5.97 | RDGNN-S | 2024-06-16 |
| Graph Neural Reaction Diffusion Models | | 93.51 ± 5.93 | RDGNN-I | 2024-06-16 |
| The Heterophilic Snowflake Hypothesis: Training and Empowering GNNs for Heterophilic Graphs | ✓ Link | 93.09 | MGNN + Hetero-S (8 layers) | 2024-06-18 |
| Higher-order Graph Convolutional Network with Flower-Petals Laplacians on Simplicial Complexes | ✓ Link | 92.45±0.73 | 2-HiGCN | 2023-09-22 |
| Diffusion-Jump GNNs: Homophiliation via Learnable Metric Filters | ✓ Link | 92.43±3.15 | DJ-GNN | 2023-06-29 |
| Sign is Not a Remedy: Multiset-to-Multiset Message Passing for Learning on Heterophilic Graphs | ✓ Link | 89.19 ± 4.5 | M2M-GNN | 2024-05-31 |
| Revisiting Heterophily For Graph Neural Networks | ✓ Link | 88.38 ± 3.64 | ACM-GCN+ | 2022-10-14 |
| Revisiting Heterophily For Graph Neural Networks | ✓ Link | 88.38 ± 3.43 | ACM-GCN++ | 2022-10-14 |
| Revisiting Heterophily For Graph Neural Networks | ✓ Link | 88.38 ± 3.43 | ACMII-GCN++ | 2022-10-14 |
| Graph Neural Aggregation-diffusion with Metastability | | 88.3±3.5 | GRADE-GAT | 2024-03-29 |
| Revisiting Heterophily For Graph Neural Networks | ✓ Link | 88.11 ± 3.24 | ACMII-GCN+ | 2022-10-14 |
| Revisiting Heterophily For Graph Neural Networks | ✓ Link | 87.84 ± 4.4 | ACM-GCN | 2022-10-14 |
| GCNH: A Simple Method For Representation Learning On Heterophilous Graphs | ✓ Link | 87.84±3.87 | GCNH | 2023-04-21 |
| Simple Truncated SVD based Model for Node Classification on Heterophilic Graphs | | 87.57 ± 5.44 | HLP Concat | 2021-06-24 |
| Improving Graph Neural Networks with Simple Architecture Design | ✓ Link | 87.30 ± 5.55 | FSGNN | 2021-05-17 |
| Revisiting Heterophily For Graph Neural Networks | ✓ Link | 86.76 ± 4.75 | ACMII-GCN | 2022-10-14 |
| GraphRARE: Reinforcement Learning Enhanced Graph Neural Network with Relative Entropy | | 86.76±5.80 | H2GCN-RARE (λ=1.0) | 2023-12-15 |
| Transitivity-Preserving Graph Representation Learning for Bridging Local Connectivity and Role-based Similarity | ✓ Link | 86.67 ±8.31 | UGT | 2023-08-18 |
| UniGAP: A Universal and Adaptive Graph Upsampling Approach to Mitigate Over-Smoothing in Node Classification Tasks | ✓ Link | 86.52 ± 4.8 | GraphSAGE + UniGAP | 2024-07-28 |
| Self-attention Dual Embedding for Graphs with Heterophily | | 86.49±5.12 | SADE-GCN | 2023-05-28 |
| Refining Latent Homophilic Structures over Heterophilic Graphs for Robust Graph Convolution Networks | | 86.32±4.5 | LHS | 2023-12-27 |
| Ordered GNN: Ordering Message Passing to Deal with Heterophily and Over-smoothing | ✓ Link | 86.22±4.12 | Ordered GNN | 2023-02-03 |
| Learn from Heterophily: Heterophilous Information-enhanced Graph Neural Network | ✓ Link | 86.22 ± 4.67 | HiGNN | 2024-03-26 |
| Sheaf Neural Networks with Connection Laplacians | ✓ Link | 86.16±2.24 | Conn-NSD | 2022-06-17 |
| Neural Sheaf Diffusion: A Topological Perspective on Heterophily and Oversmoothing in GNNs | ✓ Link | 85.95 ± 5.51 | O(d)-NSD | 2022-02-09 |
| Enhancing Intra-class Information Extraction for Heterophilous Graphs: One Neural Architecture Search Approach | | 85.84±4.23 | IIE-GNN | 2022-11-20 |
| CN-Motifs Perceptive Graph Neural Networks | | 85.68±5.28 | CNMPGNN | 2021-11-15 |
| Neural Sheaf Diffusion: A Topological Perspective on Heterophily and Oversmoothing in GNNs | ✓ Link | 85.67 ± 6.95 | Diag-NSD | 2022-02-09 |
| Non-Local Graph Neural Networks | ✓ Link | 85.4 ± 3.8 | NLMLP | 2020-05-29 |
| UniG-Encoder: A Universal Feature Encoder for Graph and Hypergraph Node Classification | ✓ Link | 85.40±5.3 | UniG-Encoder | 2023-08-03 |
| Two Sides of the Same Coin: Heterophily and Oversmoothing in Graph Convolutional Neural Networks | ✓ Link | 84.86 ± 4.55 | GGCN | 2021-02-12 |
| Make Heterophily Graphs Better Fit GNN: A Graph Rewiring Approach | | 84.86±5.01 | H2GCN+DHGR | 2022-09-17 |
| Transfer Entropy in Graph Convolutional Neural Networks | ✓ Link | 84.86 ± 4.55 | TE-GCNN | 2024-06-08 |
| Universal Deep GNNs: Rethinking Residual Connection in GNNs from a Path Decomposition Perspective for Preventing the Over-smoothing | | 84.60±5.32 | UDGNN (GCN) | 2022-05-30 |
| Improving Graph Neural Networks by Learning Continuous Edge Directions | ✓ Link | 84.59±4.53 | CoED | 2024-10-18 |
| Finding Global Homophily in Graph Neural Networks When Meeting Heterophily | ✓ Link | 84.32±4.15 | GloGNN | 2022-05-15 |
| Beyond Homophily with Graph Echo State Networks | | 84.3±4.4 | Graph ESN | 2022-10-27 |
| Finding Global Homophily in Graph Neural Networks When Meeting Heterophily | ✓ Link | 84.05±4.90 | GloGNN++ | 2022-05-15 |
| Bregman Graph Neural Network | ✓ Link | 84.05 ± 5.47 | ChebNet+Bregman | 2023-09-12 |
| Breaking the Entanglement of Homophily and Heterophily in Semi-supervised Node Classification | | 83.8±2.7 | ADPA | 2023-12-07 |
| Breaking the Limit of Graph Neural Networks by Improving the Assortativity of Graphs with Local Mixing Patterns | ✓ Link | 83.62 ± 5.50 | WRGAT | 2021-06-11 |
| Beyond Homophily in Graph Neural Networks: Current Limitations and Effective Designs | ✓ Link | 83.24 ± 7.07 | H2GCN-1 | 2020-06-20 |
| Tree Decomposed Graph Neural Network | ✓ Link | 83.00 ± 4.50 | TDGNN-w | 2021-08-25 |
| CAT: A Causally Graph Attention Network for Trimming Heterophilic Graph | ✓ Link | 83.0±2.5 | CATv3-sup | 2023-12-14 |
| Neural Sheaf Diffusion: A Topological Perspective on Heterophily and Oversmoothing in GNNs | ✓ Link | 82.97 ± 5.13 | Gen-NSD | 2022-02-09 |
| Revisiting Heterophily For Graph Neural Networks | ✓ Link | 81.89 ± 4.53 | ACM-SGC-1 | 2022-10-14 |
| Revisiting Heterophily For Graph Neural Networks | ✓ Link | 81.89 ± 4.53 | ACM-SGC-2 | 2022-10-14 |
| Adaptive Universal Generalized PageRank Graph Neural Network | ✓ Link | 81.35 ± 5.32 | GPRGCN | 2020-06-14 |
| FDGATII : Fast Dynamic Graph Attention with Initial Residual and Identity Mapping | ✓ Link | 80.5405 | FDGATII | 2021-10-21 |
| Beyond Homophily in Graph Neural Networks: Current Limitations and Effective Designs | ✓ Link | 80.00 ± 6.77 | H2GCN-2 | 2020-06-20 |
| MixHop: Higher-Order Graph Convolutional Architectures via Sparsified Neighborhood Mixing | ✓ Link | 77.84 ± 7.73 | MixHop | 2019-04-30 |
| Simple and Deep Graph Convolutional Networks | ✓ Link | 77.57 ± 3.83 | GCNII | 2020-07-04 |
| Beyond Low-frequency Information in Graph Convolutional Networks | ✓ Link | 76.49 ± 2.87 | FAGCN | 2021-01-04 |
| Large Scale Learning on Non-Homophilous Graphs: New Benchmarks and Strong Simple Methods | ✓ Link | 74.60 ± 8.37 | LINKX | 2021-10-27 |
| DeltaGNN: Graph Neural Network with Information Flow Control | ✓ Link | 74.05±3.08 | DeltaGNN constant | 2025-01-10 |
| Geom-GCN: Geometric Graph Convolutional Networks | ✓ Link | 67.57 | Geom-GCN-P | 2020-02-13 |
| Non-Local Graph Neural Networks | ✓ Link | 65.5 ± 6.6 | NLGCN | 2020-05-29 |
| Understanding over-squashing and bottlenecks on graphs via curvature | ✓ Link | 64.46±0.38 | SDRF | 2021-11-29 |
| Non-Local Graph Neural Networks | ✓ Link | 62.6 ± 7.1 | NLGAT | 2020-05-29 |
| Geom-GCN: Geometric Graph Convolutional Networks | ✓ Link | 59.73 | Geom-GCN-S | 2020-02-13 |
| Geom-GCN: Geometric Graph Convolutional Networks | ✓ Link | 57.58 | Geom-GCN-I | 2020-02-13 |
| Heterophilic Graph Neural Networks Optimization with Causal Message-passing | | 57.36±0.60 | LINKX+CausalMP | 2024-11-21 |