List your company

Question Answering on HotpotQA PoisonedRAG (test)

17Abstention Rate (ASR)BiRD
13.8834.945677.06May 19, 2026

Evaluation results

MethodLinks
BiRDLLM Backbone=Mistral-7B, Retrieval Model=Contriever2026.051768
BiRDLLM Backbone=Mistral-7B, Retrieval Model=ANCE2026.051867
BiRDLLM Backbone=Qwen2.5-7B, Retrieval Model=ANCE2026.052259
BiRDLLM Backbone=Qwen2.5-7B, Retrieval Model=Contriever2026.052469
BiRDLLM Backbone=Mistral-7B, Retrieval Model=DPR2026.052859
RobustRAGLLM Backbone=Mistral-7B, Retrieval Model=Contriever2026.052910
ReliabilityRAGLLM Backbone=Mistral-7B, Retrieval Model=ANCE2026.053312
BiRDLLM Backbone=Llama-3.1-8B, Retrieval Model=ANCE2026.053382
BiRDLLM Backbone=Llama-3.1-8B, Retrieval Model=Contriever2026.053385
RobustRAGLLM Backbone=Mistral-7B, Retrieval Model=DPR2026.053512
ReliabilityRAGLLM Backbone=Mistral-7B, Retrieval Model=DPR2026.053511
ReliabilityRAGLLM Backbone=Mistral-7B, Retrieval Model=Contriever2026.053613
ReliabilityRAGLLM Backbone=Llama-3.1-8B, Retrieval Model=ANCE2026.053630
ReliabilityRAGLLM Backbone=Llama-3.1-8B, Retrieval Model=Contriever2026.053829
BiRDLLM Backbone=Qwen2.5-7B, Retrieval Model=DPR2026.054162
RobustRAGLLM Backbone=Mistral-7B, Retrieval Model=ANCE2026.054111
RobustRAGLLM Backbone=Llama-3.1-8B, Retrieval Model=ANCE2026.054437
ReliabilityRAGLLM Backbone=Llama-3.1-8B, Retrieval Model=DPR2026.054426
RobustRAGLLM Backbone=Llama-3.1-8B, Retrieval Model=Contriever2026.054537
RobustRAGLLM Backbone=Llama-3.1-8B, Retrieval Model=DPR2026.055138
BiRDLLM Backbone=Llama-3.1-8B, Retrieval Model=DPR2026.055178
ReliabilityRAGLLM Backbone=Qwen2.5-7B, Retrieval Model=DPR2026.056031
InstructRAGLLM Backbone=Llama-3.1-8B, Retrieval Model=Contriever2026.056151
ReliabilityRAGLLM Backbone=Qwen2.5-7B, Retrieval Model=ANCE2026.056536
InstructRAGLLM Backbone=Llama-3.1-8B, Retrieval Model=DPR2026.056841
ReliabilityRAGLLM Backbone=Qwen2.5-7B, Retrieval Model=Contriever2026.057039
RobustRAGLLM Backbone=Qwen2.5-7B, Retrieval Model=DPR2026.057830
InstructRAGLLM Backbone=Mistral-7B, Retrieval Model=ANCE2026.057824
RobustRAGLLM Backbone=Qwen2.5-7B, Retrieval Model=Contriever2026.057941
InstructRAGLLM Backbone=Llama-3.1-8B, Retrieval Model=ANCE2026.057942
VanillaRAGLLM Backbone=Llama-3.1-8B, Retrieval Model=Contriever2026.057971
InstructRAGLLM Backbone=Mistral-7B, Retrieval Model=Contriever2026.058028
InstructRAGLLM Backbone=Qwen2.5-7B, Retrieval Model=ANCE2026.058120
InstructRAGLLM Backbone=Qwen2.5-7B, Retrieval Model=Contriever2026.058216
InstructRAGLLM Backbone=Qwen2.5-7B, Retrieval Model=DPR2026.058419
RobustRAGLLM Backbone=Qwen2.5-7B, Retrieval Model=ANCE2026.058833
VanillaRAGLLM Backbone=Llama-3.1-8B, Retrieval Model=ANCE2026.058865
InstructRAGLLM Backbone=Mistral-7B, Retrieval Model=DPR2026.058919
VanillaRAGLLM Backbone=Llama-3.1-8B, Retrieval Model=DPR2026.059061
VanillaRAGLLM Backbone=Qwen2.5-7B, Retrieval Model=ANCE2026.059324
VanillaRAGLLM Backbone=Qwen2.5-7B, Retrieval Model=Contriever2026.059329
VanillaRAGLLM Backbone=Mistral-7B, Retrieval Model=ANCE2026.059315
VanillaRAGLLM Backbone=Mistral-7B, Retrieval Model=Contriever2026.059316
VanillaRAGLLM Backbone=Qwen2.5-7B, Retrieval Model=DPR2026.059524
VanillaRAGLLM Backbone=Mistral-7B, Retrieval Model=DPR2026.059513