Scientific Discovery on TMC
93.25Solution QualityPIEVO
Evaluation Results
| Method | Links | |
|---|---|---|
| PIEVOBackbone=Gemini-2.5-Flash, Thinking Mode=Non-Thinking2026.02 | 93.25 | |
| PiFlowBackbone=Gemini-2.5-Flash, Thinking Mode=Non-Thinking2026.02 | 87.45 | |
| The AI Scientist v2Backbone=Gemini-2.5-Flash, Thinking Mode=Non-Thinking2026.02 | 84.86 | |
| The AI Scientist v1Backbone=Gemini-2.5-Flash, Thinking Mode=Non-Thinking2026.02 | 84.14 | |
| AI ResearcherBackbone=Gemini-2.5-Flash, Thinking Mode=Non-Thinking2026.02 | 83.79 | |
| PIEVOBackbone=Qwen3-32B, Thinking Mode=Non-Thinking2026.02 | 80.49 | |
| Vanilla MASBackbone=Gemini-2.5-Flash, Thinking Mode=Non-Thinking2026.02 | 79.36 | |
| ReActBackbone=Gemini-2.5-Flash, Thinking Mode=Non-Thinking2026.02 | 78.71 | |
| AI ResearcherBackbone=Qwen3-32B, Thinking Mode=Non-Thinking2026.02 | 74.17 | |
| The AI Scientist v1Backbone=Qwen3-32B, Thinking Mode=Non-Thinking2026.02 | 72.36 | |
| The AI Scientist v2Backbone=Qwen3-32B, Thinking Mode=Non-Thinking2026.02 | 71.45 | |
| PiFlowBackbone=Qwen3-32B, Thinking Mode=Non-Thinking2026.02 | 70.35 | |
| Vanilla MASBackbone=Qwen3-32B, Thinking Mode=Non-Thinking2026.02 | 63.98 | |
| ReActBackbone=Qwen3-32B, Thinking Mode=Non-Thinking2026.02 | 58.07 |