Question Answering on SCQA Q2
90PrecisionHelicase
Evaluation results
| Method | Links | |||
|---|---|---|---|---|
| HelicaseWeb Search capability=true2026.05 | 90 | 86 | 85 | |
| Claude Opus 4.6Web Search capability=false2026.05 | 75 | 48 | 55 | |
| ReActBackbone=Qwen3-235B, Web Search capability=true2026.05 | 68 | 34 | 39 | |
| DeepSeek-V3.2Web Search capability=false2026.05 | 60 | 47 | 48 | |
| Qwen3-235BWeb Search capability=false2026.05 | 52 | 44 | 44 | |
| ToTBackbone=Qwen3-235B, Web Search capability=true2026.05 | 50 | 28 | 30 | |
| GLM-5Web Search capability=false2026.05 | 44 | 47 | 40 |