Sequential Decision Making on WebShop (post-training evaluation)
14.8Rollout LengthStarting policy
Evaluation Results
| Method | Links | |
|---|---|---|
| Starting policyBackbone=Qwen2.5-7B-Instruct2026.06 | 14.8 | |
| GRPOBackbone=Qwen2.5-7B-Instruct, Training Protocol=GRPO2026.06 | 8 | |
| TRIAGEBackbone=Qwen2.5-7B-Instruct, Training Protocol=TRIAGE2026.06 | 6.82 |