Research map / Graph learning / Graph self-supervised learning
OGB-LSC: A Large-Scale Challenge for Machine Learning on Graphs
Read the paperCheck your idea against 108,967 papers
AI-extracted summary. Written by AI from the paper's title only (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| consider heterogeneous relations (R-GAT) instead of homogeneous paper graph | GNN with neighbor sampling · MAG240M | accuracy: up to 3 percentage points | Helped |
| attention aggregation instead of mean aggregation | GraphSAGE · MAG240M | accuracy | Helped |
| Concat encoder (shallow + RoBERTa) instead of Shallow or RoBERTa | TransE/ComplEx · WikiKG90M | MRR | Helped |
| remove candidate tails (v2 full-entity ranking) | ComplEx-Concat · WikiKG90M | MRR: 0.8637 -> 0.1833 | Hurt |
| add virtual node and use GIN instead of GCN | GNN · PCQM4M | MAE | Helped |
| train on only 10% of data (ablation) | GIN-virtual · PCQM4M | MAE | Hurt |
| increase model size (ablation) | GNNs · PCQM4M | MAE | Helped |
| GNN instead of MLP over Morgan fingerprint | PCQM4M baseline · PCQM4M | MAE | Helped |
| use target labels as input (label propagation) and temporal info | KDD Cup winners · MAG240M | accuracy: 3.7% | Helped |
Closest papers by meaning
- No Metric to Rule Them All: Toward Principled Evaluations of Graph-Learning Datasets (ICML 2025 · similarity 0.88)
- RDB2G-Bench: A Comprehensive Benchmark for Automatic Graph Modeling of Relational Databases (NeurIPS 2025 · similarity 0.88)
- Graph Classification via Reference Distribution Learning: Theory and Practice (NeurIPS 2024 · similarity 0.87)
- IGL-Bench: Establishing the Comprehensive Benchmark for Imbalanced Graph Learning (ICLR 2025 · similarity 0.87)
- Can Large Language Models Analyze Graphs like Professionals? A Benchmark, Datasets and Models (NeurIPS 2024 · similarity 0.87)
- Core Knowledge Learning Framework for Graph (AAAI 2025 · similarity 0.87)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.