Research map / Graph learning / Knowledge and temporal graphs
Beyond Explicit Edges: Robust Reasoning over Noisy and Sparse Knowledge Graphs
Hang Gao, Dimitris N. Metaxas
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 similarity-based expansion to LLM navigator (ablation) | INSES (GLM-4) · MuSiQue | EM: 0.32 -> 0.44 | Improved |
| add similarity-based expansion to LLM navigator (ablation) | INSES (GLM-4) · 2WikiMultiHopQA | EM: 0.57 -> 0.63 | Improved |
| add similarity-based expansion to LLM navigator (ablation) | INSES (GLM-4) · HotpotQA | EM: 0.53 -> 0.58 | Improved |
| add router (Naive RAG for simple queries) (ablation) | INSES (GLM-4) · HotpotQA | LLM Judge: 0.69 -> 0.80 | Improved |
| add router (Naive RAG for simple queries) (ablation) | INSES (GLM-4) · 2WikiMultiHopQA | LLM Judge: 0.61 -> 0.71 | Improved |
| use LLM Navigator alone on top of retrieval (ablation) | INSES (GLM-4) · HotpotQA | EM: 0.62 -> 0.53 | Hurt |
| switch backbone from GLM-4 to GPT-4o (ablation) | INSES + Router · MuSiQue | EM: 0.46 -> 0.48 | No clear effect |
| switch backbone from GLM-4 to GPT-4o (ablation) | INSES + Router · HotpotQA | EM: 0.68 -> 0.68 | No clear effect |
| use INSES instead of GraphRAG native retriever on its KG | INSES · MINE | mean accuracy: 0.478 -> 0.569 | Improved |
| use INSES instead of OpenIE native retriever on its KG | INSES · MINE | mean accuracy: 0.294 -> 0.569 | Improved |
| use graph-based RAG (GraphRAG Top-5) instead of Naive RAG | GraphRAG · HotpotQA | EM: 0.62 -> 0.43 | Hurt |
Closest papers by meaning
- Graph-constrained Reasoning: Faithful Reasoning on Knowledge Graphs with Large Language Models (ICML 2025 · similarity 0.88)
- Topology of Reasoning: Understanding Large Reasoning Models through Reasoning Graph Properties (NeurIPS 2025 · similarity 0.87)
- Towards Graph Foundation Models: Training on Knowledge Graphs Enables Transferability to General Graphs (NeurIPS 2025 · similarity 0.87)
- Talk like a Graph: Encoding Graphs for Large Language Models (ICLR 2024 · similarity 0.87)
- Reasoning on Graphs: Faithful and Interpretable Large Language Model Reasoning (ICLR 2024 · similarity 0.87)
- DuetGraph: Coarse-to-Fine Knowledge Graph Reasoning with Dual-Pathway Global-Local Fusion (NeurIPS 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.
