Qwen3.6-27B trained with GRPO (Dr.GRPO variant) to act as an orchestrator in a two-agent system: it decomposes a task, delegates subtasks to a second model, and integrates the results. Only orchestrator tokens receive gradient (train_roles: ["orch"]) — the sub-agent is never trained.

196 held-out tasks × 3 seeds = 588 runs per setting. All rows share identical settings (n=3, base_seed=42, max_turns=200, orch_style=pure); the comparison against the base model is a paired t-test over per-task means.

The two "fixed sub" rows hold the sub-agent constant, so any difference is attributable to orchestration alone rather than to the executor improving.

Reference: DeepSeek-V4-Flash orchestrating itself scores 0.4726 in the strong-sub setting. The 1001-task checkpoints match that number exactly; the 498-task ones remain significantly below it.

The served model name must match what your client sends. The model expects the orchestration prompt format used during training.