List your company

Question Answering on HotpotQA Clean (test)

87AccuracyVanillaRAG
-1.421.5544.567.45May 19, 2026

Evaluation results

MethodLinks
VanillaRAGLLM Backbone=Llama-3.1-8B, Retrieval Model=Contriever2026.0587
InstructRAGLLM Backbone=Llama-3.1-8B, Retrieval Model=Contriever2026.0582
VanillaRAGLLM Backbone=Mistral-7B, Retrieval Model=Contriever2026.0579
VanillaRAGLLM Backbone=Llama-3.1-8B, Retrieval Model=ANCE2026.0579
InstructRAGLLM Backbone=Llama-3.1-8B, Retrieval Model=ANCE2026.0578
BiRDLLM Backbone=Llama-3.1-8B, Retrieval Model=Contriever2026.0578
VanillaRAGLLM Backbone=Llama-3.1-8B, Retrieval Model=DPR2026.0575
BiRDLLM Backbone=Llama-3.1-8B, Retrieval Model=DPR2026.0575
VanillaRAGLLM Backbone=Mistral-7B, Retrieval Model=ANCE2026.0574
InstructRAGLLM Backbone=Mistral-7B, Retrieval Model=Contriever2026.0574
VanillaRAGLLM Backbone=Mistral-7B, Retrieval Model=DPR2026.0572
BiRDLLM Backbone=Llama-3.1-8B, Retrieval Model=ANCE2026.0570
BiRDLLM Backbone=Mistral-7B, Retrieval Model=Contriever2026.0569
BiRDLLM Backbone=Mistral-7B, Retrieval Model=DPR2026.0569
InstructRAGLLM Backbone=Llama-3.1-8B, Retrieval Model=DPR2026.0569
InstructRAGLLM Backbone=Qwen2.5-7B, Retrieval Model=Contriever2026.0567
VanillaRAGLLM Backbone=Qwen2.5-7B, Retrieval Model=Contriever2026.0566
VanillaRAGLLM Backbone=Qwen2.5-7B, Retrieval Model=DPR2026.0566
InstructRAGLLM Backbone=Mistral-7B, Retrieval Model=ANCE2026.0565
VanillaRAGLLM Backbone=Qwen2.5-7B, Retrieval Model=ANCE2026.0561
BiRDLLM Backbone=Qwen2.5-7B, Retrieval Model=DPR2026.0561
BiRDLLM Backbone=Mistral-7B, Retrieval Model=ANCE2026.0561
InstructRAGLLM Backbone=Mistral-7B, Retrieval Model=DPR2026.0560
RobustRAGLLM Backbone=Qwen2.5-7B, Retrieval Model=DPR2026.0559
InstructRAGLLM Backbone=Qwen2.5-7B, Retrieval Model=DPR2026.0559
RobustRAGLLM Backbone=Qwen2.5-7B, Retrieval Model=Contriever2026.0558
BiRDLLM Backbone=Qwen2.5-7B, Retrieval Model=Contriever2026.0557
RobustRAGLLM Backbone=Qwen2.5-7B, Retrieval Model=ANCE2026.0554
InstructRAGLLM Backbone=Qwen2.5-7B, Retrieval Model=ANCE2026.0554
ReliabilityRAGLLM Backbone=Qwen2.5-7B, Retrieval Model=Contriever2026.0550
BiRDLLM Backbone=Qwen2.5-7B, Retrieval Model=ANCE2026.0549
ReliabilityRAGLLM Backbone=Qwen2.5-7B, Retrieval Model=ANCE2026.0548
ReliabilityRAGLLM Backbone=Qwen2.5-7B, Retrieval Model=DPR2026.0542
ReliabilityRAGLLM Backbone=Llama-3.1-8B, Retrieval Model=Contriever2026.0526
RobustRAGLLM Backbone=Llama-3.1-8B, Retrieval Model=DPR2026.0526
RobustRAGLLM Backbone=Llama-3.1-8B, Retrieval Model=Contriever2026.0522
ReliabilityRAGLLM Backbone=Llama-3.1-8B, Retrieval Model=DPR2026.0522
RobustRAGLLM Backbone=Llama-3.1-8B, Retrieval Model=ANCE2026.0521
ReliabilityRAGLLM Backbone=Llama-3.1-8B, Retrieval Model=ANCE2026.0521
RobustRAGLLM Backbone=Mistral-7B, Retrieval Model=ANCE2026.058
ReliabilityRAGLLM Backbone=Mistral-7B, Retrieval Model=ANCE2026.056
ReliabilityRAGLLM Backbone=Mistral-7B, Retrieval Model=Contriever2026.054
RobustRAGLLM Backbone=Mistral-7B, Retrieval Model=DPR2026.054
RobustRAGLLM Backbone=Mistral-7B, Retrieval Model=Contriever2026.052
ReliabilityRAGLLM Backbone=Mistral-7B, Retrieval Model=DPR2026.052