Biomedical Reasoning on Biomni-Eval1
76.07Overall AccuracyS1-NexusAgent
Evaluation results
| Method | Links | |||||||||||
|---|---|---|---|---|---|---|---|---|---|---|---|---|
| S1-NexusAgentBackbone=Claude-4.5-Sonnet2026.02 | 76.07 | 44 | 90 | 94 | 74 | 70 | 70 | 86 | 90 | 76.74 | 66 | |
| Biomni AgentBackbone=Biomni-R0-32B2026.02 | 66.9 | 34 | 80 | 80 | 74 | 67 | 58 | 86 | 80 | 73 | 36 | |
| S1-NexusAgentBackbone=DeepSeek-V3.22026.02 | 61.14 | 28 | 80 | 74 | 68 | 50 | 46 | 70 | 74 | 67.44 | 54 | |
| Biomni AgentBackbone=Biomni-R0-8B2026.02 | 58.8 | 30 | 80 | 70 | 58 | 63 | 56 | 78 | 70 | 35 | 48 | |
| Biomni AgentBackbone=Claude-4-Sonnet2026.02 | 55.7 | 30 | 35 | 86 | 52 | 60 | 50 | 78 | 68 | 60 | 38 | |
| Biomni AgentBackbone=GPT-52026.02 | 54.6 | 40 | 0 | 72 | 54 | 63 | 62 | 74 | 62 | 63 | 56 | |
| S1-NexusAgentBackbone=Qwen3-8B-Sci2026.02 | 42.42 | 18 | 60 | 42 | 30 | 23.3 | 50 | 72 | 72 | 20.9 | 36 | |
| Biomni AgentBackbone=Qwen3-32B2026.02 | 34.6 | 6 | 70 | 52 | 20 | 3 | 20 | 70 | 62 | 16 | 26 | |
| S1-NexusAgentBackbone=Qwen3-8B2026.02 | 33.63 | 10 | 60 | 32 | 24 | 3.3 | 16 | 76 | 74 | 7 | 34 | |
| Biomni AgentBackbone=Qwen3-8B2026.02 | 31.8 | 2 | 60 | 28 | 16 | 7 | 22 | 76 | 72 | 9 | 26 |