Research map / Large language models / Retrieval and knowledge
Debate over Mixed-knowledge: A Robust Multi-Agent Reasoning Framework for Incomplete Knowledge Graph Question Answering
Jilong Liu, Pengyang Shao, Wei Qin, Fei Liu, Yonghui Yang, Richang Hong
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 RAG agent and judge to KG-only agent (ablation) | DoM (DeepSeek-v3) · IKGWQ | Hits@1 (%): 52.5 -> 84.5 | Improved |
| add KG agent and judge to RAG-only agent (ablation) | DoM (DeepSeek-v3) · IKGWQ | Hits@1 (%): 79.0 -> 84.5 | Improved |
| add RAG agent and judge to KG-only agent (ablation) | DoM (DeepSeek-v3) · CWQ (IKG-40%) | Hits@1 (%): 58.4 -> 62.0 | Improved |
| add RAG agent to KG-only agent with complete KG (ablation) | DoM (DeepSeek-v3) · CWQ (complete KG) | Hits@1 (%): 70.6 -> 73.3 | Improved |
| use mixed KG+RAG debate instead of KG-only (ablation) | DoM (DeepSeek-v3) · IKGWQ | Hits@1 (%): 52.5 -> 84.5 | Improved |
| use DoM instead of GoG | GoG · IKGWQ | Hits@1 (%): 49.5 -> 84.5 | Improved |
| use DoM instead of GoG | GoG · CWQ (IKG-40%) | Hits@1 (%): 60.4 -> 62.0 | Improved |
| use DoM instead of GoG | GoG · WebQSP (IKG-40%) | Hits@1 (%): 78.1 -> 81.7 | Improved |
| use stronger backbone (DeepSeek-v3 vs Qwen2.5-72B) | DoM · IKGWQ | Hits@1 (%): 69.0 -> 84.5 | Improved |
| add retrieval and debate steps vs GoG | GoG · CWQ (IKG-40%) cost | Runtime (min): 52 -> 61 | Hurt |
| add KG-based retrieval to LLM-only CoT | CoT (DeepSeek-v3) · IKGWQ | Hits@1 (%): 51.0 -> 49.5 (GoG) | Hurt |
| increase KG incompleteness (remove more triples) for DoM | DoM · CWQ IKG-20% to IKG-80% | Hits@1 (%) | Hurt |
Closest papers by meaning
- Debate on Graph: A Flexible and Reliable Reasoning Framework for Large Language Models (AAAI 2025 · similarity 0.91)
- Graph Reasoning Transformers for Knowledge-Aware Question Answering (AAAI 2024 · similarity 0.90)
- Diagnosing and Addressing Pitfalls in KG-RAG Datasets: Toward More Reliable Benchmarking (NeurIPS 2025 · similarity 0.89)
- Knowledge Graph Prompting for Multi-Document Question Answering (AAAI 2024 · similarity 0.89)
- Harnessing Large Language Models for Knowledge Graph Question Answering via Adaptive Multi-Aspect Retrieval-Augmentation (AAAI 2025 · similarity 0.88)
- Knowledge Editing with Dynamic Knowledge Graphs for Multi-Hop Question Answering (AAAI 2025 · similarity 0.88)
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.
