Research map / Graph learning / Graph self-supervised learning
When LLMs Encounter Open-world Graph Learning: A Fresh View on Unlabeled Data Uncertainty
Yanzhe Wen, Xunkai Li, Qi Zhang, Lei Zhu, Guang Zeng, Zhihan Zhang, Rong-Hua Li, Guoren Wang
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (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 |
|---|---|---|---|
| add concept modeling (CM) (ablation) | OGA ALT · Cora | coverage / precision (%): 41.25 / 62.31 -> 94.87 / 82.66 | Improved |
| add concept modeling (CM) (ablation) | OGA ALT · Citeseer | coverage / precision (%): 39.24 / 19.12 -> 93.00 / 65.19 | Improved |
| add topology-aware propagation (TP) (ablation) | OGA ALT · Cora | coverage / precision (%): 73.15 / 73.52 -> 94.87 / 82.66 | Improved |
| add topology-aware propagation (TP) (ablation) | OGA ALT · Pubmed | coverage / precision (%): 69.83 / 61.27 -> 79.76 / 98.53 | Improved |
| add semantic community detection (ablation) | OGA GLA · Citeseer | annotation accuracy: 67.3% -> 72.1% | Improved |
| add semantic community detection (ablation) | OGA GLA · Pubmed | annotation accuracy: 81.2% -> 86.0% | Improved |
| use topology-guided community annotation instead of node-wise LLM calls | OGA GLA · Cora | LLM calls: 2,708 -> 329 (87.8% down) | Improved |
| retrain GCN on OGA-annotated graph | GCN · Pubmed | accuracy (%): 47.12 -> 87.15 | Improved |
| retrain GAT on OGA-annotated graph | GAT · Pubmed | accuracy (%): 46.33 -> 86.45 | Improved |
| increase lambda sharpness to a very large value (ablation) | OGA ALT · Citeseer | UCR accuracy | Hurt |
| use OGA instead of OpenWgl for UCR | OGA · arXiv | running time (s): 2223.27 -> 99.23 | Improved |
Closest papers by meaning
- Label-free Node Classification on Graphs with Large Language Models (LLMs) (ICLR 2024 · similarity 0.89)
- Generalization Principles for Inference over Text-Attributed Graphs with Large Language Models (ICML 2025 · similarity 0.89)
- LoReC: Rethinking Large Language Models for Graph Data Analysis (NeurIPS 2026 · similarity 0.89)
- LLaGA: Large Language and Graph Assistant (ICML 2024 · similarity 0.89)
- GLBench: A Comprehensive Benchmark for Graph with Large Language Models (NeurIPS 2024 · similarity 0.89)
- How Do Large Language Models Understand Graph Patterns? A Benchmark for Graph Pattern Comprehension (ICLR 2025 · similarity 0.89)
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.
