| Diffusion-Jump GNNs: Homophiliation via Learnable Metric Filters | ✓ Link | 80.48±1.46 | DJ-GNN | 2023-06-29 |
| HoloNets: Spectral Convolutions do extend to Directed Graphs | ✓ Link | 80.33±1.19 | FaberNet | 2023-10-03 |
| Edge Directionality Improves Learning on Heterophilic Graphs | ✓ Link | 79.71±1.26 | Dir-GNN | 2023-05-17 |
| Improving Graph Neural Networks by Learning Continuous Edge Directions | ✓ Link | 79.69±1.35 | CoED | 2024-10-18 |
| Improving Graph Neural Networks with Simple Architecture Design | ✓ Link | 78.27±1.28 | FSGNN (8-hop) | 2021-05-17 |
| Improving Graph Neural Networks with Simple Architecture Design | ✓ Link | 78.14±1.25 | FSGNN (3-hop) | 2021-05-17 |
| Simple Truncated SVD based Model for Node Classification on Heterophilic Graphs | | 77.48±0.80 | HLP Concat | 2021-06-24 |
| Beyond Homophily with Graph Echo State Networks | | 76.2±1.2 | Graph ESN | 2022-10-27 |
| Self-attention Dual Embedding for Graphs with Heterophily | | 75.57±1.57 | SADE-GCN | 2023-05-28 |
| Sign is Not a Remedy: Multiset-to-Multiset Message Passing for Learning on Heterophilic Graphs | ✓ Link | 75.20 ± 2.3 | M2M-GNN | 2024-05-31 |
| Graph Neural Reaction Diffusion Models | | 74.79 ± 2.14 | RDGNN-I | 2024-06-16 |
| Revisiting Heterophily For Graph Neural Networks | ✓ Link | 74.76 ± 2.2 | ACMII-GCN++ | 2022-10-14 |
| Make Heterophily Graphs Better Fit GNN: A Graph Rewiring Approach | | 74.57±2.56 | GCNII+DHGR | 2022-09-17 |
| Revisiting Heterophily For Graph Neural Networks | ✓ Link | 74.56 ± 2.08 | ACMII-GCN+ | 2022-10-14 |
| Universal Deep GNNs: Rethinking Residual Connection in GNNs from a Path Decomposition Perspective for Preventing the Over-smoothing | | 74.53±1.19 | UDGNN (GCN) | 2022-05-30 |
| Revisiting Heterophily For Graph Neural Networks | ✓ Link | 74.47 ± 1.84 | ACM-GCN+ | 2022-10-14 |
| Revisiting Heterophily For Graph Neural Networks | ✓ Link | 74.41 ± 1.49 | ACM-GCN++ | 2022-10-14 |
| Label-Wise Graph Convolutional Network for Heterophilic Graphs | ✓ Link | 74.4±1.4 | LW-GCN | 2021-10-15 |
| SignGT: Signed Attention-based Graph Transformer for Graph Representation Learning | | 74.31±1.24 | SignGT | 2023-10-17 |
| CN-Motifs Perceptive Graph Neural Networks | | 73.29±1.29 | CNMPGNN | 2021-11-15 |
| Refining Latent Homophilic Structures over Heterophilic Graphs for Robust Graph Convolution Networks | | 72.31±1.6 | LHS | 2023-12-27 |
| Ordered GNN: Ordering Message Passing to Deal with Heterophily and Over-smoothing | ✓ Link | 72.28±2.29 | Ordered GNN | 2023-02-03 |
| Enhancing Intra-class Information Extraction for Heterophilous Graphs: One Neural Architecture Search Approach | | 72.13±2.11 | IIE-GNN | 2022-11-20 |
| GCNH: A Simple Method For Representation Learning On Heterophilous Graphs | ✓ Link | 71.56±1.86 | GCNH | 2023-04-21 |
| Finding Global Homophily in Graph Neural Networks When Meeting Heterophily | ✓ Link | 71.21±1.84 | GloGNN++ | 2022-05-15 |
| Two Sides of the Same Coin: Heterophily and Oversmoothing in Graph Convolutional Neural Networks | ✓ Link | 71.14 ± 1.84 | GGCN | 2021-02-12 |
| Transfer Entropy in Graph Convolutional Neural Networks | ✓ Link | 71.14 ± 1.84 | TE-GCNN | 2024-06-08 |
| The Heterophilic Snowflake Hypothesis: Training and Empowering GNNs for Heterophilic Graphs | ✓ Link | 70.18 | JKNet + Hetero-S (8 layers) | 2024-06-18 |
| Non-Local Graph Neural Networks | ✓ Link | 70.1 ± 2.9 | NLGCN | 2020-05-29 |
| CAT: A Causally Graph Attention Network for Trimming Heterophilic Graph | ✓ Link | 69.9±1.0 | CATv3-sup | 2023-12-14 |
| Finding Global Homophily in Graph Neural Networks When Meeting Heterophily | ✓ Link | 69.78±2.42 | GloGNN | 2022-05-15 |
| Transitivity-Preserving Graph Representation Learning for Bridging Local Connectivity and Role-based Similarity | ✓ Link | 69.78 ±3.21 | UGT | 2023-08-18 |
| GraphRARE: Reinforcement Learning Enhanced Graph Neural Network with Relative Entropy | | 69.28±1.90 | GraphSAGE-RARE (λ=1.0) | 2023-12-15 |
| Revisiting Heterophily For Graph Neural Networks | ✓ Link | 69.14 ± 1.91 | ACM-GCN | 2022-10-14 |
| Learn from Heterophily: Heterophilous Information-enhanced Graph Neural Network | ✓ Link | 68.86 ± 1.45 | HiGNN | 2024-03-26 |
| Neural Sheaf Diffusion: A Topological Perspective on Heterophily and Oversmoothing in GNNs | ✓ Link | 68.68 ± 1.73 | Diag-NSD | 2022-02-09 |
| Higher-order Graph Convolutional Network with Flower-Petals Laplacians on Simplicial Complexes | ✓ Link | 68.47±0.45 | 2-HiGCN | 2023-09-22 |
| Revisiting Heterophily For Graph Neural Networks | ✓ Link | 68.46 ± 1.7 | ACMII-GCN | 2022-10-14 |
| Large Scale Learning on Non-Homophilous Graphs: New Benchmarks and Strong Simple Methods | ✓ Link | 68.42 ± 1.38 | LINKX | 2021-10-27 |
| Restructuring Graph for Higher Homophily via Adaptive Spectral Clustering | | 68.4 ± 2.3 | LSC-ARMA | 2022-06-06 |
| Neural Sheaf Diffusion: A Topological Perspective on Heterophily and Oversmoothing in GNNs | ✓ Link | 68.04 ± 1.58 | O(d)-NSD | 2022-02-09 |
| Neural Sheaf Diffusion: A Topological Perspective on Heterophily and Oversmoothing in GNNs | ✓ Link | 67.93 ± 1.58 | Gen-NSD | 2022-02-09 |
| Non-Local Graph Neural Networks | ✓ Link | 65.7 ± 1.4 | NLGAT | 2020-05-29 |
| Breaking the Limit of Graph Neural Networks by Improving the Assortativity of Graphs with Local Mixing Patterns | ✓ Link | 65.24 ± 0.87 | WRGAT | 2021-06-11 |
| Sheaf Neural Networks with Connection Laplacians | ✓ Link | 65.21±2.04 | Conn-NSD | 2022-06-17 |
| FDGATII : Fast Dynamic Graph Attention with Initial Residual and Identity Mapping | ✓ Link | 65.1754 | FDGATII | 2021-10-21 |
| Revisiting Heterophily For Graph Neural Networks | ✓ Link | 63.99 ± 1.66 | ACM-SGC-1 | 2022-10-14 |
| Simple and Deep Graph Convolutional Networks | ✓ Link | 63.86 ± 3.04 | GCNII | 2020-07-04 |
| Adaptive Universal Generalized PageRank Graph Neural Network | ✓ Link | 62.59 ± 2.04 | GPRGCN | 2020-06-14 |
| Geom-GCN: Geometric Graph Convolutional Networks | ✓ Link | 60.9 | Geom-GCN-P | 2020-02-13 |
| MixHop: Higher-Order Graph Convolutional Architectures via Sparsified Neighborhood Mixing | ✓ Link | 60.50 ± 2.53 | MixHop | 2019-04-30 |
| Geom-GCN: Geometric Graph Convolutional Networks | ✓ Link | 60.31 | Geom-GCN-I | 2020-02-13 |
| Geom-GCN: Geometric Graph Convolutional Networks | ✓ Link | 59.96 | Geom-GCN-S | 2020-02-13 |
| Revisiting Heterophily For Graph Neural Networks | ✓ Link | 59.21 ± 2.22 | ACM-SGC-2 | 2022-10-14 |
| Heterophilic Graph Neural Networks Optimization with Causal Message-passing | | 59.14±2.42 | Gprompt+CausalMP | 2024-11-21 |
| Beyond Homophily in Graph Neural Networks: Current Limitations and Effective Designs | ✓ Link | 58.38 ± 1.76 | H2GCN-2 | 2020-06-20 |
| Beyond Homophily in Graph Neural Networks: Current Limitations and Effective Designs | ✓ Link | 52.96 ± 2.09 | H2GCN-1 | 2020-06-20 |
| Non-Local Graph Neural Networks | ✓ Link | 50.7 ± 2.2 | NLMLP | 2020-05-29 |
| Breaking the Entanglement of Homophily and Heterophily in Semi-supervised Node Classification | | 46.2±1.3 | ADPA | 2023-12-07 |
| Beyond Low-frequency Information in Graph Convolutional Networks | ✓ Link | 46.07 ± 2.11 | FAGCN | 2021-01-04 |
| Understanding over-squashing and bottlenecks on graphs via curvature | ✓ Link | 42.73±0.15 | SDRF | 2021-11-29 |