Knowledge Base Question Answering on CWQ Freebase (test)
86Hits@1ChatKBQA
Evaluation Results
| Method | Links | ||||
|---|---|---|---|---|---|
| ChatKBQAOracle entity linking=true2023.10 | 86 | 81.3 | 76.8 | — | |
| ChatKBQA2023.10 | 82.7 | 77.8 | 73.3 | — | |
| CBR-KBQA2023.10 | 70.4 | 70 | 67.1 | — | |
| DecAF2023.10 | 70.4 | — | — | — | |
| ToGOracle entity linking=true2023.10 | 69.5 | — | — | — | |
| HGNetOracle entity linking=true2023.10 | 68.9 | 68.5 | 57.8 | — | |
| KG-Hopper w/Qwen-2.5-7BSize=7B, Strategy=KG-Augmented with RL Fine-Tuning2026.03 | 61.07 | — | — | — | |
| Program TransferOracle entity linking=true2023.10 | 58.1 | 58.7 | — | — | |
| GPT-4o-mini + KGStrategy=KG-Augmented without Fine-Tuning2026.03 | 54.35 | — | — | — | |
| DeepSeek-R1-Distill-Llama-70B + KGSize=70B, Strategy=KG-Augmented without Fine-Tuning2026.03 | 52.38 | — | — | — | |
| Qwen-2.5-7B (SFT) + KGSize=7B, Strategy=KG-Augmented with Supervised Fine-Tuning2026.03 | 51.84 | — | — | — | |
| KG-CoT w/GPT 3.5-TurboStrategy=KG-Augmented without Fine-Tuning2026.03 | 51.6 | — | — | — | |
| UniKGQAOracle entity linking=true2023.10 | 51.2 | 49.4 | — | — | |
| Subgraph RetrievalOracle entity linking=true2023.10 | 50.2 | 47.1 | — | — | |
| KG-CoT w/LLaMA-13BSize=13B, Strategy=KG-Augmented with Supervised Fine-Tuning2026.03 | 50 | — | — | — | |
| NSM+hOracle entity linking=true2023.10 | 48.8 | 44 | — | — | |
| RigelOracle entity linking=true2023.10 | 48.7 | — | — | — | |
| TransferNet2023.10 | 48.6 | — | — | — | |
| LLaMA-3.1-8B (SFT) + KGSize=8B, Strategy=KG-Augmented with Supervised Fine-Tuning2026.03 | 47.4 | — | — | — | |
| PullNet2023.10 | 47.2 | — | — | — | |
| KG-CoT w/LLaMA-7BSize=7B, Strategy=KG-Augmented with Supervised Fine-Tuning2026.03 | 46.7 | — | — | — | |
| LLaMA-3.1-8B + KGSize=8B, Strategy=KG-Augmented without Fine-Tuning2026.03 | 45.64 | — | — | — | |
| Qwen-2.5-7B + KGSize=7B, Strategy=KG-Augmented without Fine-Tuning2026.03 | 44.82 | — | — | — | |
| EmbedKGQAOracle entity linking=true2023.10 | 44.7 | — | — | — | |
| QGG2023.10 | 44.1 | 40.4 | — | — | |
| LLaMA-3.3-70B + KGSize=70B, Strategy=KG-Augmented without Fine-Tuning2026.03 | 44 | — | — | — | |
| Interactive-KBQA w/LLaMA-13BSize=13B, Strategy=KG-Augmented with Supervised Fine-Tuning2026.03 | 42.5 | — | — | — | |
| GPT-4o-miniStrategy=LLM Prompting Only2026.03 | 42.32 | — | — | — | |
| GPT-4oStrategy=LLM Prompting Only2026.03 | 41.77 | — | — | — | |
| TextRay2023.10 | 40.8 | 33.9 | — | — | |
| Interactive-KBQA w/LLaMA-7BSize=7B, Strategy=KG-Augmented with Supervised Fine-Tuning2026.03 | 39.9 | — | — | — | |
| Topic Units2023.10 | 39.3 | 36.5 | — | — | |
| LLaMA-3.3-70BSize=70B, Strategy=LLM Prompting Only2026.03 | 37.2 | — | — | — | |
| GRAFT-Net2023.10 | 36.8 | 32.7 | — | — | |
| LLaMA-3.1-8BSize=8B, Strategy=LLM Prompting Only2026.03 | 32.33 | — | — | — | |
| DeepSeek-R1-Distill-Llama-70BSize=70B, Strategy=LLM Prompting Only2026.03 | 31.92 | — | — | — | |
| Qwen-2.5-7BSize=7B, Strategy=LLM Prompting Only2026.03 | 31.25 | — | — | — | |
| KV-Mem2023.10 | 21.1 | 15.7 | — | — | |
| Chain-of-ThoughtBackbone=GPT-4o2025.10 | — | — | — | 52.4 | |
| Debate-on-GraphBackbone=GPT-42025.10 | — | — | — | 56 | |
| EffiQABackbone=GPT-42025.10 | — | — | — | 69.5 | |
| FC-KBQA2023.10 | — | 56.4 | — | — | |
| FiDeLiSBackbone=GPT-4-Turbo2025.10 | — | — | — | 71.5 | |
| FiDeLiSBackbone=GPT-4o2025.10 | — | — | — | 65.3 | |
| Generate-on-GraphBackbone=GPT-42025.10 | — | — | — | 75.2 | |
| GMT-KBQA2023.10 | — | 77 | 72.2 | — | |
| GRAFT-Net2025.10 | — | — | — | 32.8 | |
| IO PromptBackbone=GPT-4o2025.10 | — | — | — | 51.7 | |
| iQUESTBackbone=GPT-4o2025.10 | — | — | — | 73.9 | |
| KELDaRBackbone=GPT-42025.10 | — | — | — | 63 | |
| ORTBackbone=GPT-4o2025.10 | — | — | — | 65.4 | |
| Plan-on-GraphBackbone=GPT-42025.10 | — | — | — | 75 | |
| PullNet2025.10 | — | — | — | 47.2 | |
| ReadiBackbone=GPT-42025.10 | — | — | — | 67 | |
| ReKnoSBackbone=GPT-42025.10 | — | — | — | 68.2 | |
| Self-ConsistencyBackbone=GPT-4o2025.10 | — | — | — | 53.8 | |
| SoGBackbone=GPT-4o2025.10 | — | — | — | 75.1 | |
| SubgraphRAGBackbone=GPT-4o2025.10 | — | — | — | 67.5 | |
| Think-on-GraphBackbone=GPT-42025.10 | — | — | — | 69.5 | |
| Think-on-GraphBackbone=GPT-4o, Reproduced status=reproduced2025.10 | — | — | — | 48.9 | |
| TransferNet2025.10 | — | — | — | 48.6 | |
| UHop2023.10 | — | 29.8 | — | — | |
| UnifiedSKG2023.10 | — | 68.8 | — | — | |
| UniKGQA2025.10 | — | — | — | 51.2 |