Task-free exploration on WA Shopping
43.1Success RateAMZ + Target + WA
Evaluation Results
| Method | Links | ||
|---|---|---|---|
| AMZ + Target + WATraining Setting=Self-guided Exploration, Iterations=1502025.10 | 43.1 | 14.6 | |
| WATraining Setting=Single-Domain Specialists, Iterations=502025.10 | 42.3 | 14.9 | |
| WA → Target → AMZTraining Setting=Sequential Curriculum, Iterations=50 + 50 + 502025.10 | 42.1 | 10.8 | |
| AMZ → WATraining Setting=Sequential Curriculum, Iterations=75 + 752025.10 | 40.2 | 12.3 | |
| SkillWeaver*Training Setting=Sequential Curriculum, Iterations=1502025.10 | 39.8 | 8.6 | |
| Target → AMZ → WATraining Setting=Sequential Curriculum, Iterations=50 + 50 + 502025.10 | 39.5 | 11.5 | |
| AMZ → Target → WATraining Setting=Sequential Curriculum, Iterations=50 + 50 + 502025.10 | 38.2 | 11.9 | |
| AMZTraining Setting=Single-Domain Specialists, Iterations=502025.10 | 38.1 | 2.7 | |
| TargetTraining Setting=Single-Domain Specialists, Iterations=502025.10 | 38 | 2.1 | |
| BaselineIterations=–2025.10 | 37.4 | — |