Research Idea Generation on Expert Evaluation Biotech Domain (test)
91.7Novelty Winrate vs Qwen-14BDeepInnovator
Evaluation Results
| Method | Links | ||||||||
|---|---|---|---|---|---|---|---|---|---|
| DeepInnovatorevaluation_mode=Human Expert Annotation2026.02 | 91.7 | 61.5 | 100 | 38.5 | 58.3 | 53.8 | 90.9 | 35.7 |
Loading the SOTA2 catalog…