Scientific Foundation Model Collaboration on EywaBench Space
71.87Utility ScoreEywaOrchestra
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| EywaOrchestraSetting=Dynamic Orchestration2026.04 | 71.87 | 75.47 | 10,810 | |
| EywaMASSetting=Multi-Agent2026.04 | 68.99 | 72.51 | 11,787 | |
| Debate MASSetting=Multi-Agent2026.04 | 65.43 | 74.75 | 13,614 | |
| Refine MASSetting=Multi-Agent2026.04 | 62.44 | 79.65 | 10,043 | |
| EywaAgentSetting=Single-Agent2026.04 | 61.23 | 23.12 | 3,329 | |
| MoASetting=Multi-Agent2026.04 | 58.63 | 69.32 | 17,332 | |
| X-MASSetting=Multi-Agent2026.04 | 57.23 | 79.06 | 16,584 | |
| Single-LLM-AgentSetting=Single-Agent2026.04 | 52.35 | 26 | 4,512 |