Research map / Graph learning / Graph self-supervised learning
GraphPrompt: Graph-Based Prompt Templates for Biomedical Synonym Prediction
Hanwen Xu, Jiayou Zhang, Zhirui Wang, Shizhuo Zhang, Megh Bhalerao, Yucong Liu, Dawei Zhu, Sheng 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 |
|---|---|---|---|
| Use graph-based prompt templates with first- and second-order neighbors instead of text-only BioBERT representations | GraphPrompt · mp zero-shot | Acc@1: 76.47 -> 81.14 | Improved |
| Use graph-based prompt templates with first- and second-order neighbors instead of text-only BioBERT representations | GraphPrompt · mp few-shot | Acc@1: 85.78 -> 88.86 | Improved |
| Use graph-based prompt templates with first- and second-order neighbors instead of text-only BioBERT representations | GraphPrompt · cl zero-shot | Acc@1: 65.14 -> 67.81 | Improved |
| Use graph-based prompt templates with first- and second-order neighbors instead of text-only BioBERT representations | GraphPrompt · cl few-shot | Acc@1: 76.64 -> 82.86 | Improved |
| Use graph-based prompt templates with first- and second-order neighbors instead of text-only BioBERT representations | GraphPrompt · hp zero-shot | Acc@1: 61.53 -> 69.14 | Improved |
| Use graph-based prompt templates with first- and second-order neighbors instead of text-only BioBERT representations | GraphPrompt · hp few-shot | Acc@1: 72.04 -> 77.87 | Improved |
| Use graph-based prompt templates with first- and second-order neighbors instead of text-only BioBERT representations | GraphPrompt · fbbt zero-shot | Acc@1: 65.38 -> 69.01 | Improved |
| Use graph-based prompt templates with first- and second-order neighbors instead of text-only BioBERT representations | GraphPrompt · fbbt few-shot | Acc@1: 69.76 -> 75.55 | Improved |
| Use graph-based prompt templates with first- and second-order neighbors instead of text-only BioBERT representations | GraphPrompt · doid zero-shot | Acc@1: 50.71 -> 59.06 | Improved |
| Use graph-based prompt templates with first- and second-order neighbors instead of text-only BioBERT representations | GraphPrompt · doid few-shot | Acc@1: 59.40 -> 67.53 | Improved |
| Add second-order graph templates (ablation) | GraphPrompt without T² · cl zero-shot | Acc@1: 68.00 -> 67.81 | Hurt |
| Add second-order graph templates (ablation) | GraphPrompt without T² · cl few-shot | Acc@1: 82.04 -> 82.86 | Improved |
Later papers that compared with this method
8 later accepted papers report GraphPrompt as a baseline in their results tables; 7 of them report a better number on most of the results they share with it.
- SA²GFM: Enhancing Robust Graph Foundation Models with Structure-Aware Semantic Augmentation (AAAI 2026 · better on 1 of 1 shared result)
- GraphTOP: Graph Topology-Oriented Prompting for Graph Neural Networks (NeurIPS 2025 · better on 3 of 5 shared results)
- HeterGP: Bridging Heterogeneity in Graph Neural Networks with Multi-View Prompting (AAAI 2025 · better on 2 of 4 shared results)
- Node-Time Conditional Prompt Learning in Dynamic Graphs (ICLR 2025 · better on 3 of 3 shared results)
- Towards Graph Foundation Models: Learning Generalities Across Graphs via Task-Trees (ICML 2025 · better on 1 of 1 shared result)
- One Prompt Fits All: Universal Graph Adaptation for Pretrained Models (NeurIPS 2025 · better on 1 of 1 shared result)
- HGPrompt: Bridging Homogeneous and Heterogeneous Graphs for Few-Shot Prompt Learning (AAAI 2024 · better on 2 of 2 shared results)
- Universal Prompt Tuning for Graph Neural Networks (NeurIPS 2023 · better on 1 of 1 shared result)
Matched by method name in results tables read by AI. Authors' own tables tend to favour the newer method, and a comparison can be missed when a method is renamed.
Closest papers by meaning
- A Lightweight Neural Model for Biomedical Entity Linking (AAAI 2021 · similarity 0.86)
- Can Embeddings Adequately Represent Medical Terminology? New Large-Scale Medical Term Similarity Datasets Have the Answer! (AAAI 2020 · similarity 0.85)
- LATTE: Latent Type Modeling for Biomedical Entity Linking (AAAI 2020 · similarity 0.84)
- Reasoning-Enhanced Healthcare Predictions with Knowledge Graph Community Retrieval (ICLR 2025 · similarity 0.84)
- GraphCare: Enhancing Healthcare Predictions with Personalized Knowledge Graphs (ICLR 2024 · similarity 0.83)
- Seeing Unseen: Discover Novel Biomedical Concepts via Geometry-Constrained Probabilistic Modeling (CVPR 2024 · similarity 0.81)
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.
