| Clarify Confused Nodes via Separated Learning | ✓ Link | 43.89 ± 1.33 | NCSAGE | 2023-06-04 |
| Clarify Confused Nodes via Separated Learning | ✓ Link | 43.16 ± 1.32 | NCGCN | 2023-06-04 |
| Higher-order Graph Convolutional Network with Flower-Petals Laplacians on Simplicial Complexes | ✓ Link | 41.81±0.52 | 2-HiGCN | 2023-09-22 |
| Enhancing Intra-class Information Extraction for Heterophilous Graphs: One Neural Architecture Search Approach | | 39.91 ± 2.41 | IIE-GNN | 2022-11-20 |
| Refining Latent Homophilic Structures over Heterophilic Graphs for Robust Graph Convolution Networks | | 38.87±1.0 | LHS | 2023-12-27 |
| Graph Neural Reaction Diffusion Models | | 38.69 ± 1.41 | RDGNN-I | 2024-06-16 |
| SignGT: Signed Attention-based Graph Transformer for Graph Representation Learning | | 38.65±0.32 | SignGT | 2023-10-17 |
| CAT: A Causally Graph Attention Network for Trimming Heterophilic Graph | ✓ Link | 38.5±1.2 | CATv3-sup | 2023-12-14 |
| Ordered GNN: Ordering Message Passing to Deal with Heterophily and Over-smoothing | ✓ Link | 37.99 ± 1.00 | Ordered GNN | 2023-02-03 |
| Mamba-Based Graph Convolutional Networks: Tackling Over-smoothing with Selective State Space | | 37.97±0.91 | MbaGCN | 2025-01-26 |
| Mixture of Experts Meets Decoupled Message Passing: Towards General and Adaptive Node Classification | ✓ Link | 37.97±1.01 | GNNMoE(SAGE-like P) | 2024-12-11 |
| Self-attention Dual Embedding for Graphs with Heterophily | | 37.91 ± 0.97 | SADE-GCN | 2023-05-28 |
| Non-Local Graph Neural Networks | ✓ Link | 37.9 ± 1.3 | NLMLP | 2020-05-29 |
| Neural Sheaf Diffusion: A Topological Perspective on Heterophily and Oversmoothing in GNNs | ✓ Link | 37.81 ± 1.15 | O(d)-NSD | 2022-02-09 |
| Neural Sheaf Diffusion: A Topological Perspective on Heterophily and Oversmoothing in GNNs | ✓ Link | 37.80 ± 1.22 | Gen-NSD | 2022-02-09 |
| Neural Sheaf Diffusion: A Topological Perspective on Heterophily and Oversmoothing in GNNs | ✓ Link | 37.79 ± 1.01 | Diag-NSD | 2022-02-09 |
| Mixture of Experts Meets Decoupled Message Passing: Towards General and Adaptive Node Classification | ✓ Link | 37.76±0.98 | GNNMoE(GAT-like P) | 2024-12-11 |
| Finding Global Homophily in Graph Neural Networks When Meeting Heterophily | ✓ Link | 37.7 ± 1.40 | GloGNN++ | 2022-05-15 |
| UniGAP: A Universal and Adaptive Graph Upsampling Approach to Mitigate Over-Smoothing in Node Classification Tasks | ✓ Link | 37.69 ± 1.2 | GGCN + UniGAP | 2024-07-28 |
| Mixture of Experts Meets Decoupled Message Passing: Towards General and Adaptive Node Classification | ✓ Link | 37.59±1.36 | GNNMoE(GCN-like P) | 2024-12-11 |
| Two Sides of the Same Coin: Heterophily and Oversmoothing in Graph Convolutional Neural Networks | ✓ Link | 37.54 ± 1.56 | GGCN | 2021-02-12 |
| Transfer Entropy in Graph Convolutional Neural Networks | ✓ Link | 37.50±1.57 | TE-GCNN | 2024-06-08 |
| Simple and Deep Graph Convolutional Networks | ✓ Link | 37.44 ± 1.30 | GCNII | 2020-07-04 |
| Make Heterophily Graphs Better Fit GNN: A Graph Rewiring Approach | | 37.43 ± 0.78 | GPRGNN+DHGR | 2022-09-17 |
| Finding Global Homophily in Graph Neural Networks When Meeting Heterophily | ✓ Link | 37.35 ± 1.30 | GloGNN | 2022-05-15 |
| Revisiting Heterophily For Graph Neural Networks | ✓ Link | 37.31 ± 1.09 | ACM-GCN++ | 2022-10-14 |
| Heterophilous Distribution Propagation for Graph Neural Networks | | 37.26 ± 0.67 | HDP | 2024-05-31 |
| Learn from Heterophily: Heterophilous Information-enhanced Graph Neural Network | ✓ Link | 37.21 ± 1.35 | HiGNN | 2024-03-26 |
| Revisiting Heterophily For Graph Neural Networks | ✓ Link | 37.09 ± 1.32 | ACMII-GCN++ | 2022-10-14 |
| Diffusion-Jump GNNs: Homophiliation via Learnable Metric Filters | ✓ Link | 36.93 ± 0.84 | DJ-GNN | 2023-06-29 |
| GCNH: A Simple Method For Representation Learning On Heterophilous Graphs | ✓ Link | 36.89 ± 1.50 | GCNH | 2023-04-21 |
| Sign is Not a Remedy: Multiset-to-Multiset Message Passing for Learning on Heterophilic Graphs | ✓ Link | 36.72 ± 1.6 | M2M-GNN | 2024-05-31 |
| Revisiting Heterophily For Graph Neural Networks | ✓ Link | 36.63 ± 0.84 | ACM-GCN | 2022-10-14 |
| Breaking the Limit of Graph Neural Networks by Improving the Assortativity of Graphs with Local Mixing Patterns | ✓ Link | 36.53 ± 0.77 | WRGAT | 2021-06-11 |
| Revisiting Heterophily For Graph Neural Networks | ✓ Link | 36.31 ± 1.2 | ACMII-GCN | 2022-10-14 |
| Revisiting Heterophily For Graph Neural Networks | ✓ Link | 36.26 ± 1.34 | ACM-GCN+ | 2022-10-14 |
| CN-Motifs Perceptive Graph Neural Networks | | 36.25 ± 0.98 | CNMPGNN | 2021-11-15 |
| Restructuring Graph for Higher Homophily via Adaptive Spectral Clustering | | 36.2 ± 1.0 | LSC-GNN | 2022-06-06 |
| Revisiting Heterophily For Graph Neural Networks | ✓ Link | 36.14 ± 1.44 | ACMII-GCN+ | 2022-10-14 |
| Universal Deep GNNs: Rethinking Residual Connection in GNNs from a Path Decomposition Perspective for Preventing the Over-smoothing | | 36.13 ± 1.21 | UDGNN (GCN) | 2022-05-30 |
| Large Scale Learning on Non-Homophilous Graphs: New Benchmarks and Strong Simple Methods | ✓ Link | 36.10 ± 1.55 | LINKX | 2021-10-27 |
| Revisiting Heterophily For Graph Neural Networks | ✓ Link | 36.04 ± 0.83 | ACM-SGC-2 | 2022-10-14 |
| The Heterophilic Snowflake Hypothesis: Training and Empowering GNNs for Heterophilic Graphs | ✓ Link | 35.99 | MGNN + Hetero-S (4 layers) | 2024-06-18 |
| Bregman Graph Neural Network | ✓ Link | 35.92 ± 0.84 | ChebNet+Bregman | 2023-09-12 |
| Improving Graph Neural Networks with Simple Architecture Design | ✓ Link | 35.75 ± 0.96 | FSGNN (8-hop) | 2021-05-17 |
| Revisiting Heterophily For Graph Neural Networks | ✓ Link | 35.49 ± 1.06 | ACM-SGC-1 | 2022-10-14 |
| Adaptive Universal Generalized PageRank Graph Neural Network | ✓ Link | 35.16 ± 0.9 | GPRGCN | 2020-06-14 |
| Beyond Low-frequency Information in Graph Convolutional Networks | ✓ Link | 34.82 ± 1.35 | FAGCN | 2021-01-04 |
| Simple Truncated SVD based Model for Node Classification on Heterophilic Graphs | | 34.59 ± 1.32 | HLP Concat | 2021-06-24 |
| Addressing Heterophily in Node Classification with Graph Echo State Networks | ✓ Link | 34.56 ± 0.76 | GESN | 2023-05-14 |
| Beyond Homophily with Graph Echo State Networks | | 34.5 ± 0.8 | Graph ESN | 2022-10-27 |
| Beyond Homophily in Graph Neural Networks: Current Limitations and Effective Designs | ✓ Link | 34.49 ± 1.63 | H2GCN-2 | 2020-06-20 |
| Beyond Homophily in Graph Neural Networks: Current Limitations and Effective Designs | ✓ Link | 34.31 ± 1.31 | H2GCN-1 | 2020-06-20 |
| MixHop: Higher-Order Graph Convolutional Architectures via Sparsified Neighborhood Mixing | ✓ Link | 32.22 ± 2.34 | MixHop | 2019-04-30 |
| DiffWire: Inductive Graph Rewiring via the Lovász Bound | ✓ Link | 31.98 | CT-Layer | 2022-06-15 |
| Geom-GCN: Geometric Graph Convolutional Networks | ✓ Link | 31.63 | Geom-GCN-P | 2020-02-13 |
| Non-Local Graph Neural Networks | ✓ Link | 31.6 ± 1.0 | NLGCN | 2020-05-29 |
| Geom-GCN: Geometric Graph Convolutional Networks | ✓ Link | 30.3 | Geom-GCN-S | 2020-02-13 |
| Non-Local Graph Neural Networks | ✓ Link | 29.5 ± 1.3 | NLGAT | 2020-05-29 |
| DiffWire: Inductive Graph Rewiring via the Lovász Bound | ✓ Link | 29.35 | CT-Layer (PE) | 2022-06-15 |
| Geom-GCN: Geometric Graph Convolutional Networks | ✓ Link | 29.09 | Geom-GCN-I | 2020-02-13 |
| Understanding over-squashing and bottlenecks on graphs via curvature | ✓ Link | 28.42 ± 0.75 | SDRF | 2021-11-29 |