Question Answering on OpenBookQA (dev)
90AccuracyKnowGPT
Evaluation results
| Method | Links | |
|---|---|---|
| KnowGPTCategory=Ours2023.12 | 90 | |
| GPT-4Category=LLM + Zero-shot2023.12 | 87.8 | |
| JointLKCategory=KG-enhanced LM2023.12 | 86.4 | |
| HamQACategory=KG-enhanced LM2023.12 | 85.8 | |
| GreaseLMCategory=KG-enhanced LM2023.12 | 85.7 | |
| MindmapCategory=LLM + KG Prompting2023.12 | 85.1 | |
| GrapeQACategory=KG-enhanced LM2023.12 | 84.9 | |
| QA-GNNCategory=KG-enhanced LM2023.12 | 83.6 | |
| CoKCategory=LLM + KG Prompting2023.12 | 83.5 | |
| RoGCategory=LLM + KG Prompting2023.12 | 82.3 | |
| MHGRNCategory=KG-enhanced LM2023.12 | 78.6 | |
| Llama3Category=LLM + Zero-shot, Parameters=8b2023.12 | 77.1 | |
| RoBerta-largeCategory=LM + Fine tuning2023.12 | 66.8 | |
| Bert-largeCategory=LM + Fine tuning2023.12 | 62.6 | |
| GPT-3.5Category=LLM + Zero-shot, Implementation=gpt-3.5-turbo2023.12 | 59.8 | |
| Bert-baseCategory=LM + Fine tuning2023.12 | 58.8 | |
| Llama2Category=LLM + Zero-shot, Parameters=7b2023.12 | 52.4 | |
| GPT-3Category=LLM + Zero-shot, Implementation=text-davinci-0022023.12 | 42 | |
| BaichuanCategory=LLM + Zero-shot, Parameters=7B2023.12 | 41.1 | |
| ChatGLM2Category=LLM + Zero-shot2023.12 | 39.2 | |
| InternLMCategory=LLM + Zero-shot2023.12 | 37.6 | |
| ChatGLMCategory=LLM + Zero-shot2023.12 | 35.2 |