Market Trading Prediction on PolyBench
80.9AccuracyFull System
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| Full SystemSystem Category=Ours2026.06 | 80.9 | 330 | |
| Multi agentSystem Category=Ours2026.06 | 79.8 | 351 | |
| AdaptiveSystem Category=Ours2026.06 | 77.4 | 352 | |
| Meta HarnessSystem Category=Auto-harness baselines2026.06 | 50.8 | 320 | |
| OctoToolsSystem Category=Human2026.06 | 40 | 20.4 | |
| SonnetSystem Category=No evolution2026.06 | 22.2 | 1.7 | |
| SkillOSSystem Category=Auto-harness baselines2026.06 | 21.4 | 3.6 | |
| A-EvolveSystem Category=Auto-harness baselines2026.06 | 18.4 | 7.2 | |
| HaikuSystem Category=No evolution2026.06 | 15.2 | 88 | |
| KimiSystem Category=No evolution2026.06 | 14 | -2 | |
| GLMSystem Category=No evolution2026.06 | 14 | 10.1 | |
| GEPASystem Category=Auto-harness baselines2026.06 | 13.4 | 0.2 | |
| Cont. HarnessSystem Category=Auto-harness baselines2026.06 | 8.5 | 1.7 | |
| DeepSeekSystem Category=No evolution2026.06 | 1.4 | 16.4 |