mlsanityData to 14 Oct 2026

Self-Supervised Continual Graph Learning in Adaptive Riemannian Spaces

arXiv 2211.17068 ↗First posted 30 Nov 2022cs.LGcited by 32

On the map: Knowledge distillation → Graph neural network distillation

Abstract (arXiv)

Continual graph learning routinely finds its role in a variety of real-world applications where the graph data with different tasks come sequentially. Despite the success of prior works, it still faces great challenges. On the one hand, existing methods work with the zero-curvature Euclidean space, and largely ignore the fact that curvature varies over the coming graph sequence. On the other hand, continual learners in the literature rely on abundant labels, but labeling graph in practice is particularly hard especially for the continuously emerging graphs on-the-fly. To address the aforementioned challenges, we propose to explore a challenging yet practical problem, the self-supervised continual graph learning in adaptive Riemannian spaces. In this paper, we propose a novel self-supervised Riemannian Graph Continual Learner (RieGrace). In RieGrace, we first design an Adaptive Riemannian GCN (AdaRGCN), a unified GCN coupled with a neural curvature adapter, so that Riemannian space is shaped by the learnt curvature adaptive to each graph. Then, we present a Label-free Lorentz Distillation approach, in which we create teacher-student AdaRGCN for the graph sequence. The student successively performs intra-distillation from itself and inter-distillation from the teacher so as to consolidate knowledge without catastrophic forgetting. In particular, we propose a theoretically grounded Generalized Lorentz Projection for the contrastive distillation in Riemannian space. Extensive experiments on the benchmark datasets show the superiority of RieGrace, and additionally, we investigate on how curvature changes over the graph sequence.

Key facts. Cited by 32 papers in the map (data through 14 Oct 2026). Placed in Graph neural network distillation, one of 12 directions in Knowledge distillation. Builds on 42 of its 58 references that are in the map. First posted on arXiv 30 Nov 2022.

Builds on

42 of 58 references.

Closest ancestors · 1

Same direction, newest first

  1. 2022Geometer: Graph Few-Shot Class-Incremental Learning via Prototype Representation27 May

Other references · 41

  1. 2016Semi-Supervised Classification with Graph Convolutional Networkscited by 11,361
  2. 2017Graph Attention Networkscited by 8,923
  3. 2018Representation Learning with Contrastive Predictive Codingcited by 7,777
  4. 2017Inductive Representation Learning on Large Graphscited by 6,320
  5. 2013Distributed Representations of Words and Phrases and their Compositionalitycited by 3,474
  6. 2016Learning without Forgettingcited by 2,496
  7. 2017Gradient Episodic Memory for Continual Learningcited by 1,898
  8. 2019Simplifying Graph Convolutional Networkscited by 1,477
  9. 2018Deep Graph Infomaxcited by 1,106
  10. 2019Contrastive Representation Distillationcited by 748
  11. 2017Poincar'e Embeddings for Learning Hierarchical Representationscited by 684
  12. 2020Contrastive Multi-View Representation Learning on Graphscited by 652
29 more
  1. 2020GCC: Graph Contrastive Coding for Graph Neural Network Pre-Trainingcited by 463
  2. 2019Hyperbolic Graph Convolutional Neural Networkscited by 402
  3. 2019Hyperbolic Graph Neural Networkscited by 244
  4. 2018Representation Tradeoffs for Hyperbolic Embeddingscited by 206
  5. 2018Hyperbolic Attention Networkscited by 122
  6. 2019Continuous Hierarchical Representations with Poincar'e Variational Auto-Encoderscited by 112
  7. 2019Constant Curvature Graph Convolutional Networkscited by 108
  8. 2021Fully Hyperbolic Neural Networkscited by 108
  9. 2020Overcoming Catastrophic Forgetting in Graph Neural Networkscited by 95
  10. 2021Achieving Forgetting Prevention and Knowledge Transfer in Continual Learningcited by 91
  11. 2019Mixed-curvature Variational Autoencoderscited by 86
  12. 2021AutoGCL: Automated Graph Contrastive Learning via Learnable View Generatorscited by 73
  13. 2021Lorentzian Graph Convolutional Networkscited by 68
  14. 2021Discrete-time Temporal Network Embedding via Implicit Hierarchical Learning in Hyperbolic Spacecited by 67
  15. 2020Streaming Graph Neural Networks via Continual Learningcited by 64
  16. 2021Hyperbolic Variational Graph Neural Network for Modeling Dynamic Graphscited by 50
  17. 2020GraphSAIL: Graph Structure Aware Incremental Learning for Recommender Systemscited by 42
  18. 2021TrafficStream: A Streaming Traffic Flow Forecasting Framework Based on Graph Neural Networks and Continual Learningcited by 42
  19. 2021A Self-supervised Mixed-curvature Graph Neural Networkcited by 33
  20. 2022Graph Lifelong Learning: A Surveycited by 30
  21. 2020Computationally Tractable Riemannian Manifolds for Graph Embeddingscited by 29
  22. 2021Hierarchical Prototype Networks for Continual Graph Representation Learningcited by 28
  23. 2022A Self-supervised Riemannian GNN with Time Varying Curvature for Temporal Graph Learningcited by 24
  24. 2020Lifelong Learning of Graph Neural Networks for Open-World Node Classificationcited by 22
  25. 2016Characterizing Complex Networks with Forman-Ricci Curvature and Associated Geometric Flowscited by 21
  26. 2019Hyperbolic Disk Embeddings for Directed Acyclic Graphscited by 21
  27. 2020SAIL: Self-Augmented Graph Contrastive Learningcited by 20
  28. 2022Dual Space Graph Contrastive Learningcited by 20
  29. 2022Ultrahyperbolic Knowledge Graph Embeddingscited by 20

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2023: 4; 2024: 15; 2025: 10; 2026: 3.

32 papers.

Sort:

By area:

  1. 2024Continual Learning with Pre-Trained Models: A Surveycited by 154
  2. 2025RiemannGFM: Learning a Graph Foundation Model from Riemannian Geometrycited by 35
  3. 2024Motif-aware Riemannian Graph Neural Network with Generative-Contrastive Learningcited by 23
  4. 2024Continual Learning on Graphs: Challenges, Solutions, and Opportunitiescited by 21
  5. 2024LSEnet: Lorentz Structural Entropy Neural Network for Deep Graph Clusteringcited by 21
  6. 2023Continual Graph Learning: A Surveycited by 18
  7. 2024Graph Learning under Distribution Shifts: A Comprehensive Survey on Domain Adaptation, Out-of-distribution, and Continual Learningcited by 18
  8. 2025Position: Beyond Euclidean -- Foundation Models Should Embrace Non-Euclidean Geometriescited by 18
  9. 2024DeepRicci: Self-supervised Graph Structure-Feature Co-Refinement for Alleviating Over-squashingcited by 16
  10. 2024Spiking Graph Neural Network on Riemannian Manifoldscited by 16
  11. 2023Contrastive Graph Clustering in Curvature Spacescited by 15
  12. 2024Topology-aware Embedding Memory for Continual Learning on Expanding Networkscited by 14
20 more
  1. 2024Continual Learning on Graphs: A Surveycited by 12
  2. 2024Replay-and-Forget-Free Graph Class-Incremental Learning: A Task Profiling and Prompting Approachcited by 12
  3. 2025GCAL: Adapting Graph Models to Evolving Domain Shiftscited by 8
  4. 2023Poincar'e Differential Privacy for Hierarchy-Aware Graph Embeddingcited by 7
  5. 2024Towards Continuous Reuse of Graph Models via Holistic Memory Diversificationcited by 7
  6. 2025Deeper with Riemannian Geometry: Overcoming Oversmoothing and Oversquashing for Graph Foundation Modelscited by 7
  7. 2023SINCERE: Sequential Interaction Networks representation learning on Co-Evolving RiEmannian manifoldscited by 4
  8. 2025ASIL: Augmented Structural Information Learning for Deep Graph Clustering in Hyperbolic Spacecited by 4
  9. 2026DisRFM: Polar Riemannian Flow Matching for Structure-Preserving Graph Domain Adaptationcited by 4
  10. 2024Contrastive Sequential Interaction Network Learning on Co-Evolving Riemannian Spacescited by 3
  11. 2024RCoCo: Contrastive Collective Link Prediction across Multiplex Network in Riemannian Spacecited by 2
  12. 2024R-ODE: Ricci Curvature Tells When You Will be Informedcited by 2
  13. 2025Dynamic Mixture-of-Experts for Incremental Graph Learningcited by 1
  14. 2025The Unreasonable Effectiveness of Randomized Representations in Online Continual Graph Learningcited by 1
  15. 2026Riemannian Geometry Speaks Louder Than Words: From Graph Foundation Model to Next-Generation Graph Intelligencecited by 1
  16. 2024Alleviating Over-Smoothing via Aggregation over Compact Manifoldscited by 0
  17. 2025MEGA: Second-Order Gradient Alignment for Catastrophic Forgetting Mitigation in GFSCILcited by 0
  18. 2025Feature Staleness Aware Incremental Learning for CTR Predictioncited by 0
  19. 2025PR-CapsNet: Pseudo-Riemannian Capsule Network with Adaptive Curvature Routing for Graph Learningcited by 0
  20. 2026SAOT: Self-Supervised Continual Graph Learning with Structure-Aware Optimal Transportcited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2025Toward a Unified Geometry Understanding: Riemannian Diffusion Framework for Graph Generation and Prediction
  2. 2026Geometric Mixture-of-Experts with Curvature-Guided Adaptive Routing for Graph Representation Learning
  3. 2026Graph Representation via Elements of Discrete Morse and Cobordism Theories
  4. 2025Task-Aware Adaptive Modulation: A Replay-Free and Resource-Efficient Approach For Continual Graph Learning
  5. 2023Dynamic Sub-graph Distillation for Robust Semi-supervised Continual Learning
  6. 2026Distilling Graph Geometry: Knowledge Gap from GNNs to MLPs
  7. 2022Spectral Maps for Learning on Subgraphs
  8. 2026Learning from Demonstrations over Riemannian Manifolds using Neural ODEs: An Extended Abstract ↗