DynaFlow: Learning Free-Form Multi-Agent Orchestration
Paper · https://arxiv.org/abs/2610.34467
Multi-agent systems often rely on predefined collaboration structures, limiting their ability to adapt to diverse task requirements. We introduce DynaFlow, a framework that integrates free-form executable workflow synthesis into the action space of the main agent. The agent decides when collaboration is needed and generates task-specific programs that coordinate sub-agents through sequential or parallel execution, structured information exchange, and verification. Execution results and traces inform subsequent actions, enabling the agent to adapt collaboration as reasoning progresses. We train only the main agent through supervised fine-tuning on distilled trajectories, followed by refinement on examples filtered for unsupported premises, biased verification, and information-flow defects. All sub-agents remain fixed, separating orchestration learning from executor capability improvement. Across six benchmarks spanning mathematical reasoning, scientific question answering, and multi-hop retrieval, DynaFlow achieves 88.08% overall accuracy in Forced mode, outperforming the strongest evaluated baseline by 10.90 percentage points with the same Qwen3.5-27B backbone. Workflow analysis finds more explicit use, verification, and refinement of upstream outputs after training, highlighting cross-stage information flow as a key dimension of multi-agent orchestration.
First seen by Oriel on .
AI summary
DynaFlow trains a main agent to synthesize adaptive, executable multi-agent workflows, reaching 88.08% accuracy across six reasoning and retrieval benchmarks.
- The main agent decides when collaboration is needed and generates programs coordinating fixed sub-agents through sequential or parallel execution, structured information exchange, and verification.
- Execution results and traces guide later actions, allowing collaboration workflows to change as reasoning progresses.
- Training combines supervised fine-tuning on distilled trajectories with refinement using examples filtered for unsupported premises, biased verification, and information-flow defects.
- In Forced mode with a Qwen3.5-27B backbone, DynaFlow outperforms the strongest evaluated baseline by 10.90 percentage points.
- Post-training workflows more explicitly use, verify, and refine upstream outputs, identifying cross-stage information flow as important to orchestration.
#multi-agent systems #llm agents #workflow synthesis #supervised fine-tuning
Generated by gpt-5.6-sol. Check the original for details.
Metric trends
Last 30 days · Up to 120 recorded samples per source and metric, with reposts kept separate · First and latest retained
Points
Score reported by the source.
- Hugging Face Papers: 39, latest observation ; -141 over observed period since
Comments
Comment count reported by the source.
- Hugging Face Papers: 3, latest observation ; +2 over observed period since
Aliases
| Source | External ID | URL |
|---|---|---|
| Hugging Face Papers | 2610.34467 | https://arxiv.org/abs/2610.34467 Read discussion |