List your company

Multi-hop QA on 2WikiMHQA Wiki18 index (test)

55.1Pass@4Qwen3-8B
23.79631.92340.0548.177Jun 15, 2026

Evaluation results

MethodLinks
Qwen3-8BSampling Method=DivInit (DI)2026.0655.1
Gemma3-12BSampling Method=DivInit (DI)2026.0653.9
Gemma3-4BSampling Method=DivInit (DI)2026.0652.2
Gemma3-12BSampling Method=Standard parallel sampling (S)2026.0652
Gemma3-12BSampling Method=DivInit (DI)2026.0650.2
Qwen3-4BSampling Method=DivInit (DI)2026.0649
Qwen3-8BSampling Method=Standard parallel sampling (S)2026.0646.3
Qwen3-8BSampling Method=DivInit (DI)2026.0646
Gemma3-12BSampling Method=Standard parallel sampling (S)2026.0645
Gemma3-4BSampling Method=Standard parallel sampling (S)2026.0642.8
Qwen3-4BSampling Method=Standard parallel sampling (S)2026.0641.9
Qwen3-1.7BSampling Method=DivInit (DI)2026.0641.5
Qwen3-8BSampling Method=Standard parallel sampling (S)2026.0638.6
Qwen3-1.7BSampling Method=Standard parallel sampling (S)2026.0637.6
Qwen3-4BSampling Method=DivInit (DI)2026.0636.6
Gemma3-4BSampling Method=DivInit (DI)2026.0634
Qwen3-4BSampling Method=Standard parallel sampling (S)2026.0629.5
Gemma3-4BSampling Method=Standard parallel sampling (S)2026.0628
Qwen3-1.7BSampling Method=DivInit (DI)2026.0627.8
Qwen3-1.7BSampling Method=Standard parallel sampling (S)2026.0625