Question Answering on HotpotQA Clean (test)
87AccuracyVanillaRAG
Evaluation Results
| Method | Links | |
|---|---|---|
| VanillaRAGLLM Backbone=Llama-3.1-8B, Retrieval Model=Contriever2026.05 | 87 | |
| InstructRAGLLM Backbone=Llama-3.1-8B, Retrieval Model=Contriever2026.05 | 82 | |
| VanillaRAGLLM Backbone=Mistral-7B, Retrieval Model=Contriever2026.05 | 79 | |
| VanillaRAGLLM Backbone=Llama-3.1-8B, Retrieval Model=ANCE2026.05 | 79 | |
| InstructRAGLLM Backbone=Llama-3.1-8B, Retrieval Model=ANCE2026.05 | 78 | |
| BiRDLLM Backbone=Llama-3.1-8B, Retrieval Model=Contriever2026.05 | 78 | |
| VanillaRAGLLM Backbone=Llama-3.1-8B, Retrieval Model=DPR2026.05 | 75 | |
| BiRDLLM Backbone=Llama-3.1-8B, Retrieval Model=DPR2026.05 | 75 | |
| VanillaRAGLLM Backbone=Mistral-7B, Retrieval Model=ANCE2026.05 | 74 | |
| InstructRAGLLM Backbone=Mistral-7B, Retrieval Model=Contriever2026.05 | 74 | |
| VanillaRAGLLM Backbone=Mistral-7B, Retrieval Model=DPR2026.05 | 72 | |
| BiRDLLM Backbone=Llama-3.1-8B, Retrieval Model=ANCE2026.05 | 70 | |
| BiRDLLM Backbone=Mistral-7B, Retrieval Model=Contriever2026.05 | 69 | |
| BiRDLLM Backbone=Mistral-7B, Retrieval Model=DPR2026.05 | 69 | |
| InstructRAGLLM Backbone=Llama-3.1-8B, Retrieval Model=DPR2026.05 | 69 | |
| InstructRAGLLM Backbone=Qwen2.5-7B, Retrieval Model=Contriever2026.05 | 67 | |
| VanillaRAGLLM Backbone=Qwen2.5-7B, Retrieval Model=Contriever2026.05 | 66 | |
| VanillaRAGLLM Backbone=Qwen2.5-7B, Retrieval Model=DPR2026.05 | 66 | |
| InstructRAGLLM Backbone=Mistral-7B, Retrieval Model=ANCE2026.05 | 65 | |
| VanillaRAGLLM Backbone=Qwen2.5-7B, Retrieval Model=ANCE2026.05 | 61 | |
| BiRDLLM Backbone=Qwen2.5-7B, Retrieval Model=DPR2026.05 | 61 | |
| BiRDLLM Backbone=Mistral-7B, Retrieval Model=ANCE2026.05 | 61 | |
| InstructRAGLLM Backbone=Mistral-7B, Retrieval Model=DPR2026.05 | 60 | |
| RobustRAGLLM Backbone=Qwen2.5-7B, Retrieval Model=DPR2026.05 | 59 | |
| InstructRAGLLM Backbone=Qwen2.5-7B, Retrieval Model=DPR2026.05 | 59 | |
| RobustRAGLLM Backbone=Qwen2.5-7B, Retrieval Model=Contriever2026.05 | 58 | |
| BiRDLLM Backbone=Qwen2.5-7B, Retrieval Model=Contriever2026.05 | 57 | |
| RobustRAGLLM Backbone=Qwen2.5-7B, Retrieval Model=ANCE2026.05 | 54 | |
| InstructRAGLLM Backbone=Qwen2.5-7B, Retrieval Model=ANCE2026.05 | 54 | |
| ReliabilityRAGLLM Backbone=Qwen2.5-7B, Retrieval Model=Contriever2026.05 | 50 | |
| BiRDLLM Backbone=Qwen2.5-7B, Retrieval Model=ANCE2026.05 | 49 | |
| ReliabilityRAGLLM Backbone=Qwen2.5-7B, Retrieval Model=ANCE2026.05 | 48 | |
| ReliabilityRAGLLM Backbone=Qwen2.5-7B, Retrieval Model=DPR2026.05 | 42 | |
| ReliabilityRAGLLM Backbone=Llama-3.1-8B, Retrieval Model=Contriever2026.05 | 26 | |
| RobustRAGLLM Backbone=Llama-3.1-8B, Retrieval Model=DPR2026.05 | 26 | |
| RobustRAGLLM Backbone=Llama-3.1-8B, Retrieval Model=Contriever2026.05 | 22 | |
| ReliabilityRAGLLM Backbone=Llama-3.1-8B, Retrieval Model=DPR2026.05 | 22 | |
| RobustRAGLLM Backbone=Llama-3.1-8B, Retrieval Model=ANCE2026.05 | 21 | |
| ReliabilityRAGLLM Backbone=Llama-3.1-8B, Retrieval Model=ANCE2026.05 | 21 | |
| RobustRAGLLM Backbone=Mistral-7B, Retrieval Model=ANCE2026.05 | 8 | |
| ReliabilityRAGLLM Backbone=Mistral-7B, Retrieval Model=ANCE2026.05 | 6 | |
| ReliabilityRAGLLM Backbone=Mistral-7B, Retrieval Model=Contriever2026.05 | 4 | |
| RobustRAGLLM Backbone=Mistral-7B, Retrieval Model=DPR2026.05 | 4 | |
| RobustRAGLLM Backbone=Mistral-7B, Retrieval Model=Contriever2026.05 | 2 | |
| ReliabilityRAGLLM Backbone=Mistral-7B, Retrieval Model=DPR2026.05 | 2 |