Autonomous Scientific Research on ARC-Bench 12-topic
67.5Code Development ScoreSAGE
Evaluation results
| Method | Links | |||||
|---|---|---|---|---|---|---|
| SAGE2026.06 | 67.5 | 62.5 | 34.6 | 52 | 7 | |
| AI-Scientist-v22026.06 | 58.3 | 51.7 | 39.2 | 48.2 | 2 | |
| AutoResearchClaw2026.06 | 33.3 | 20.8 | 21.7 | 24.8 | — |
| Method | Links | |||||
|---|---|---|---|---|---|---|
| SAGE2026.06 | 67.5 | 62.5 | 34.6 | 52 | 7 | |
| AI-Scientist-v22026.06 | 58.3 | 51.7 | 39.2 | 48.2 | 2 | |
| AutoResearchClaw2026.06 | 33.3 | 20.8 | 21.7 | 24.8 | — |