Open-domain QA on Overall QA Suite (NQ, TriviaQA, PopQA, HotpotQA, 2WikiMultiHopQA, MuSiQue, Bamboogle)
62.3Average EMSDGA-Phase
Evaluation Results
| Method | Links | |
|---|---|---|
| SDGA-PhaseFramework=ZeroSearch, Model=Qwen2.5-7B, Search Depth (Smax)=52026.05 | 62.3 | |
| SDGA-AutoFramework=ZeroSearch, Model=Qwen2.5-7B, Search Depth (Smax)=52026.05 | 58.3 | |
| SDGA-PhaseFramework=ZeroSearch, Model=Qwen2.5-3B, Search Depth (Smax)=52026.05 | 56.3 | |
| SDGA-AutoFramework=ZeroSearch, Model=Qwen2.5-3B, Search Depth (Smax)=52026.05 | 52.9 | |
| GRPO-FullFramework=ZeroSearch, Model=Qwen2.5-7B2026.05 | 50.9 | |
| SDGA-PhaseFramework=Search-R1, Model=Qwen2.5-7B, Search Depth (Smax)=52026.05 | 47.1 | |
| GRPO-HalfFramework=ZeroSearch, Model=Qwen2.5-7B2026.05 | 44.7 | |
| GRPO-FullFramework=ZeroSearch, Model=Qwen2.5-3B2026.05 | 44.5 | |
| SDGA-AutoFramework=Search-R1, Model=Qwen2.5-7B, Search Depth (Smax)=52026.05 | 42.2 | |
| SDGA-PhaseFramework=Search-R1, Model=Qwen2.5-3B, Search Depth (Smax)=52026.05 | 40.9 | |
| SDGA-AutoFramework=Search-R1, Model=LLaMA-3.2-3B-Base, Search Depth (Smax)=52026.05 | 39.6 | |
| SDGA-PhaseFramework=Search-R1, Model=LLaMA-3.2-3B-Base, Search Depth (Smax)=52026.05 | 39.4 | |
| GRPO-HalfFramework=ZeroSearch, Model=Qwen2.5-3B2026.05 | 39.3 | |
| GRPO-FullFramework=Search-R1, Model=Qwen2.5-7B2026.05 | 39 | |
| SDGA-PhaseFramework=ZeroSearch, Model=LLaMA-3.2-3B-Base, Search Depth (Smax)=52026.05 | 38.8 | |
| SDGA-AutoFramework=ZeroSearch, Model=LLaMA-3.2-3B-Base, Search Depth (Smax)=52026.05 | 38.5 | |
| SDGA-AutoFramework=Search-R1, Model=Qwen2.5-3B, Search Depth (Smax)=52026.05 | 37.5 | |
| GRPO-FullFramework=Search-R1, Model=Qwen2.5-3B2026.05 | 35.9 | |
| GRPO-HalfFramework=Search-R1, Model=Qwen2.5-7B2026.05 | 34.4 | |
| GRPO-FullFramework=ZeroSearch, Model=LLaMA-3.2-3B-Base2026.05 | 33.6 | |
| GRPO-HalfFramework=Search-R1, Model=Qwen2.5-3B2026.05 | 33.2 | |
| GRPO-FullFramework=Search-R1, Model=LLaMA-3.2-3B-Base2026.05 | 31.9 | |
| GRPO-HalfFramework=ZeroSearch, Model=LLaMA-3.2-3B-Base2026.05 | 29.7 | |
| GRPO-HalfFramework=Search-R1, Model=LLaMA-3.2-3B-Base2026.05 | 28.6 | |
| SDGA-AntiFramework=ZeroSearch, Model=Qwen2.5-7B, Search Depth (Smax)=52026.05 | 22.4 | |
| SDGA-AntiFramework=Search-R1, Model=Qwen2.5-7B, Search Depth (Smax)=52026.05 | 20.5 | |
| SDGA-AntiFramework=ZeroSearch, Model=Qwen2.5-3B, Search Depth (Smax)=52026.05 | 20.4 | |
| SDGA-AntiFramework=Search-R1, Model=Qwen2.5-3B, Search Depth (Smax)=52026.05 | 18.3 | |
| SDGA-AntiFramework=ZeroSearch, Model=LLaMA-3.2-3B-Base, Search Depth (Smax)=52026.05 | 16.2 | |
| SDGA-AntiFramework=Search-R1, Model=LLaMA-3.2-3B-Base, Search Depth (Smax)=52026.05 | 15.2 |