mlsanityData to 7 Oct 2026

Simple GNN Regularisation for 3D Molecular Property Prediction & Beyond

arXiv 2106.07971 ↗First posted 15 Jun 2021cs.LGcited by 44

Not on the map. Why?

Abstract (arXiv)

In this paper we show that simple noise regularisation can be an effective way to address GNN oversmoothing. First we argue that regularisers addressing oversmoothing should both penalise node latent similarity and encourage meaningful node representations. From this observation we derive "Noisy Nodes", a simple technique in which we corrupt the input graph with noise, and add a noise correcting node-level loss. The diverse node level loss encourages latent node diversity, and the denoising objective encourages graph manifold learning. Our regulariser applies well-studied methods in simple, straightforward ways which allow even generic architectures to overcome oversmoothing and achieve state of the art results on quantum chemistry tasks, and improve results significantly on Open Graph Benchmark (OGB) datasets. Our results suggest Noisy Nodes can serve as a complementary building block in the GNN toolkit.

Key facts. Cited by 44 papers in the map (data through 7 Oct 2026). Builds on 47 of its 72 references that are in the map. First posted on arXiv 15 Jun 2021.

Builds on

47 of 72 references.

References · 47

  1. 2017Attention Is All You Needcited by 59,932
  2. 2018BERT: Pre-training of Deep Bidirectional Transformers for Language Understandingcited by 40,204
  3. 2014Adam: A Method for Stochastic Optimizationcited by 36,960
  4. 2020Language Models are Few-Shot Learnerscited by 32,199
  5. 2016Semi-Supervised Classification with Graph Convolutional Networkscited by 11,282
  6. 2018How Powerful are Graph Neural Networks?cited by 4,355
  7. 2019Generative Modeling by Estimating Gradients of the Data Distributioncited by 4,067
  8. 2017Neural Message Passing for Quantum Chemistrycited by 3,460
  9. 2019A Comprehensive Survey on Graph Neural Networkscited by 2,906
  10. 2020Open Graph Benchmark: Datasets for Machine Learning on Graphscited by 2,148
  11. 2018Graph Neural Networks: A Review of Methods and Applicationscited by 1,571
  12. 2018Relational inductive biases, deep learning, and graph networkscited by 1,341
35 more
  1. 2020Graph Contrastive Learning with Augmentationscited by 1,203
  2. 2018Deeper Insights into Graph Convolutional Networks for Semi-Supervised Learningcited by 1,099
  3. 2016Geometric deep learning: going beyond Euclidean datacited by 1,042
  4. 2019Strategies for Pre-training Graph Neural Networkscited by 964
  5. 2021E(n) Equivariant Graph Neural Networkscited by 951
  6. 2020Learning to Simulate Complex Physics with Graph Networkscited by 867
  7. 2020Learning Mesh-Based Simulation with Graph Networkscited by 840
  8. 2015Convolutional Networks on Graphs for Learning Molecular Fingerprintscited by 751
  9. 2018Tensor field networks: Rotation- and translation-equivariant neural networks for 3D point cloudscited by 735
  10. 2021E(3)-Equivariant Graph Neural Networks for Data-Efficient and Accurate Interatomic Potentialscited by 701
  11. 2020SE(3)-Transformers: 3D Roto-Translation Equivariant Attention Networkscited by 602
  12. 2017SchNet: A continuous-filter convolutional neural network for modeling quantum interactionscited by 569
  13. 2020Directional Message Passing for Molecular Graphscited by 556
  14. 2019Measuring and Relieving the Over-smoothing Problem for Graph Neural Networks from the Topological Viewcited by 544
  15. 2019DeepGCNs: Can GCNs Go as Deep as CNNs?cited by 527
  16. 2021OGB-LSC: A Large-Scale Challenge for Machine Learning on Graphscited by 328
  17. 2021Do Transformers Really Perform Bad for Graph Representation?cited by 318
  18. 2019PairNorm: Tackling Oversmoothing in GNNscited by 297
  19. 2019Cormorant: Covariant Molecular Neural Networkscited by 262
  20. 2020Fast and Uncertainty-Aware Directional Message Passing for Non-Equilibrium Moleculescited by 254
  21. 2020DeeperGCN: All You Need to Train Deeper GCNscited by 249
  22. 2020The Open Catalyst 2020 (OC20) Dataset and Community Challengescited by 246
  23. 2020A Note on Over-Smoothing for Graph Neural Networkscited by 245
  24. 2018Graph networks as learnable physics engines for inference and controlcited by 242
  25. 2020Random Features Strengthen Graph Neural Networkscited by 184
  26. 2021Training Graph Neural Networks with 1000 Layerscited by 169
  27. 2021Spherical Message Passing for 3D Graph Networkscited by 166
  28. 2019Relational Pooling for Graph Representationscited by 150
  29. 2020Building powerful and equivariant graph neural networks with structural message-passingcited by 95
  30. 2021ForceNet: A Graph Neural Network for Large-Scale Quantum Calculationscited by 58
  31. 2019Equivariant Flows: sampling configurations for multi-body systems with symmetric energiescited by 57
  32. 2020Revisiting Over-smoothing in Deep GCNscited by 55
  33. 2018Covariant Compositional Networks For Learning Graphscited by 54
  34. 2019Coloring graph neural networks for node disambiguationcited by 49
  35. 2020Graph Convolutional Neural Networks with Node Transition Probability-based Message Passing and DropNode Regularizationcited by 15

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2022: 6; 2023: 17; 2024: 12; 2025: 5; 2026: 4.

44 papers.

Sort:

By area:

  1. 2022Equiformer: Equivariant Graph Attention Transformer for 3D Atomistic Graphscited by 258
  2. 2022The CLRS Algorithmic Reasoning Benchmarkcited by 103
  3. 2022One Transformer Can Understand Both 2D & 3D Molecular Datacited by 77
  4. 2023FAENet: Frame Averaging Equivariant GNN for Materials Modelingcited by 72
  5. 2023A Hitchhiker's Guide to Geometric GNNs for 3D Atomic Systemscited by 65
  6. 2024Orb: A Fast, Scalable Neural Network Potentialcited by 63
  7. 2022Spherical Channels for Modeling Atomic Interactionscited by 47
  8. 2023Smooth, exact rotational symmetrization for deep learning on point cloudscited by 39
  9. 2023Trainable Projected Gradient Method for Robust Fine-tuningcited by 37
  10. 2024On the Scalability of GNNs for Molecular Graphscited by 36
  11. 2023Fractional Denoising for 3D Molecular Pre-trainingcited by 33
  12. 2023Evaluation of the MACE Force Field Architecture: from Medicinal Chemistry to Materials Sciencecited by 25
32 more
  1. 2022Latent Graph Inference using Product Manifoldscited by 23
  2. 2023Diffusion-based Generative AI for Exploring Transition States from 2D Molecular Graphscited by 19
  3. 2024Evaluating representation learning on the protein structure universecited by 16
  4. 2023Improving Compositional Generalization Using Iterated Learning and Simplicial Embeddingscited by 15
  5. 2023May the Force be with You: Unified Force-Centric Pre-Training for 3D Molecular Conformationscited by 13
  6. 2024Triplet Interaction Improves Graph Transformers: Accurate Molecular Graph Learning with Triplet Graph Transformerscited by 12
  7. 2023Accelerating Molecular Graph Neural Networks via Knowledge Distillationcited by 11
  8. 2023Wigner kernels: body-ordered equivariant machine learning without a basiscited by 7
  9. 2024Impact of Domain Knowledge and Multi-Modality on Intelligent Molecular Property Prediction: A Systematic Surveycited by 6
  10. 2023GPS++: Reviving the Art of Message Passing for Molecular Property Predictioncited by 5
  11. 2023GeoTMI:Predicting quantum chemical property with easy-to-obtain geometry via positional denoisingcited by 5
  12. 2023DR-Label: Improving GNN Models for Catalysis Systems by Label Deconstruction and Reconstructioncited by 4
  13. 2022Curvature-informed multi-task learning for graph networkscited by 3
  14. 2023SupSiam: Non-contrastive Auxiliary Loss for Learning from Molecular Conformerscited by 3
  15. 2023Principles for Initialization and Architecture Selection in Graph Neural Networks with ReLU Activationscited by 3
  16. 2024PASCL: Supervised Contrastive Learning with Perturbative Augmentation for Particle Decay Reconstructioncited by 3
  17. 2026Dynamics-inspired Structure Hallucination for Protein-protein Interaction Modelingcited by 3
  18. 2026Enhancing Molecular Property Predictions by Learning from Bond Modelling and Interactionscited by 3
  19. 2024Advancing Molecular Machine Learning Representations with Stereoelectronics-Infused Molecular Graphscited by 2
  20. 2025M4GN: Mesh-based Multi-segment Hierarchical Graph Network for Dynamic Simulationscited by 2
  21. 2025Scaling Kinetic Monte-Carlo Simulations of Grain Growth with Combined Convolutional and Graph Neural Networkscited by 2
  22. 2025ToxBench: A Binding Affinity Prediction Benchmark with AB-FEP-Calculated Labels for Human Estrogen Receptor Alphacited by 1
  23. 2023Generating QM1B with PySCFIPUcited by 0
  24. 2024Entropy Aware Message Passing in Graph Neural Networkscited by 0
  25. 2024DualBind: A Dual-Loss Framework for Protein-Ligand Binding Affinity Predictioncited by 0
  26. 2024Improving Molecular Modeling with Geometric GNNs: an Empirical Studycited by 0
  27. 2024Harnessing the Power of Noise: A Survey of Techniques and Applicationscited by 0
  28. 2024Conformation Generation using Transformer Flowscited by 0
  29. 2025Topological Feature Compression for Molecular Graph Neural Networkscited by 0
  30. 2025Learning 3D Anisotropic Noise Distributions Improves Molecular Force Field Modelingcited by 0
  31. 2026Enabling Structure-Only Initialization and Out-of-Distribution Generalization in GNN-based Molecular Dynamics Simulatorscited by 0
  32. 2026MEGA-CL: A Molecular Foundation Model for Generalizable ADMET Prediction through Graph External Attention and Contrastive Learningcited by 0