DynaFlow framework released
The story so far
Qwen has released the DynaFlow framework, which coordinates sub-agents by having the main agent synthesize executable processes to achieve adaptive multi-agent collaboration. The framework integrates free-form executable process synthesis into the action space of the main agent, where the main agent determines when collaboration is needed and generates programs for specific tasks, coordinating sub-agents through sequential or parallel execution, structured information exchange, and verification. Qwen fine-tunes the main agent through supervised micro-training and trains it on distilled trajectories, subsequently improving examples that filter out unsupported premises, biased validations, and information flow defects. All sub-agents remain fixed, separating orchestration learning from actuator capability improvement. On six benchmark tests covering mathematical reasoning, scientific Q&A, and multi-hop retrieval, DynaFlow achieved an overall accuracy of 88.08% in forced mode, which is 10.90 percentage points higher than the strongest evaluation baseline compared to the same Qwen3.5-27B main branch.
AI-written from the reports · Updated 23 days ago
Coverage timeline
Follow the coverage to see every side of the story.
- Hugging Face Daily PapersDynaFlow: Learning Free-Form Multi-Agent Orchestration
DynaFlow coordinates sub-agents by having the main agent synthesize executable processes, achieving adaptive multi-agent collaboration. The framework integrates free-form executable process synthesis into the action space of the main agent, where the main agent determines when collaboration is needed and generates programs for specific tasks, coordinating sub-agents through sequential or parallel execution, structured information exchange, and verification.