List your company

Evidence Retrieval on FAKEWIKI RolePlay

63.8Recall@10SCORINGMODEL
-2.44814.75131.9549.149May 7, 2026

Evaluation results

MethodLinks
SCORINGMODELTarget Model=Qwen3-8B2026.0563.8
SCORINGMODELTarget Model=Llama-3.1-8B2026.0563.4
STEERFUSETarget Model=Qwen3-8B2026.0560.5
STEERFUSETarget Model=Llama-3.1-8B2026.0543.5
BGE-baseTarget Model=Qwen3-8B, Query Type=QA2026.0540.3
BGE-baseTarget Model=Llama-3.1-8B, Query Type=QA2026.0530.6
SBERT-MiniLMTarget Model=Qwen3-8B, Query Type=QA2026.0529.1
SBERT-MPNetTarget Model=Qwen3-8B, Query Type=QA2026.0527.5
ContrieverTarget Model=Qwen3-8B, Query Type=answer2026.0526.2
BGE-baseTarget Model=Qwen3-8B, Query Type=answer2026.0526
Finetuned EmbedSimTarget Model=Llama-3.1-8B2026.0523.6
SBERT-MiniLMTarget Model=Qwen3-8B, Query Type=answer2026.0523.5
ContrieverTarget Model=Qwen3-8B, Query Type=QA2026.0522.3
SBERT-MPNetTarget Model=Llama-3.1-8B, Query Type=QA2026.0521.6
SBERT-MPNetTarget Model=Qwen3-8B, Query Type=answer2026.0520.3
SBERT-MiniLMTarget Model=Llama-3.1-8B, Query Type=QA2026.0519.9
Finetuned EmbedSimTarget Model=Qwen3-8B2026.0516.1
BGE-baseTarget Model=Llama-3.1-8B, Query Type=answer2026.0511.7
ContrieverTarget Model=Llama-3.1-8B, Query Type=answer2026.0511.3
SBERT-MiniLMTarget Model=Llama-3.1-8B, Query Type=answer2026.0510.6
SBERT-MPNetTarget Model=Llama-3.1-8B, Query Type=answer2026.059.7
ContrieverTarget Model=Llama-3.1-8B, Query Type=QA2026.059.3
MinHashTarget Model=Qwen3-8B, Query Type=answer2026.050.1
MinHashTarget Model=Qwen3-8B, Query Type=QA2026.050.1
MinHashTarget Model=Llama-3.1-8B, Query Type=answer2026.050.1
MinHashTarget Model=Llama-3.1-8B, Query Type=QA2026.050.1