Scientific Foundation Model Collaboration on EywaBench Infrastructure
74.06UtilityEywaMAS
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| EywaMASSetting=Multi-Agent2026.04 | 74.06 | 46.49 | 8,912 | |
| EywaOrchestraSetting=Dynamic Orchestration2026.04 | 72.98 | 28.8 | 6,892 | |
| Debate MASSetting=Multi-Agent2026.04 | 72.37 | 60.73 | 10,311 | |
| X-MASSetting=Multi-Agent2026.04 | 71.8 | 62.95 | 14,056 | |
| EywaAgentSetting=Single-Agent2026.04 | 70.6 | 15.99 | 2,248 | |
| MoASetting=Multi-Agent2026.04 | 69.38 | 47.34 | 12,365 | |
| Single-LLM-AgentSetting=Single-Agent2026.04 | 67.58 | 18.42 | 3,327 | |
| Refine MASSetting=Multi-Agent2026.04 | 63.62 | 47.49 | 7,438 |