Saved in:
| Main Authors: | , , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2604.25012 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866915962057392128 |
|---|---|
| author | Du, Shiyi Liu, Jiayuan Du, Weihua Huang, Yue Li, Jiayi Luo, Yingtao Zhang, Xiangliang Conitzer, Vincent Kingsford, Carl |
| author_facet | Du, Shiyi Liu, Jiayuan Du, Weihua Huang, Yue Li, Jiayi Luo, Yingtao Zhang, Xiangliang Conitzer, Vincent Kingsford, Carl |
| contents | Automated agentic workflow design currently relies on per-task iterative search, which is computationally prohibitive and fails to reuse structural knowledge across tasks. We observe that optimized workflows converge to a small family of domain-specific topologies, suggesting that this combinatorial search is largely redundant. Building on this insight, we propose SWIFT (Synthesizing Workflows via Few-shot Transfer), a framework that amortizes workflow design into reusable structural priors. SWIFT first distills compositional heuristics and output-interface contracts from contrastive analysis of prior search trajectories across source tasks. At inference time, it conditions a single LLM generation pass on these priors together with cross-task workflow demonstrations to synthesize a complete, executable workflow for an unseen target task, bypassing iterative search entirely. On five benchmarks, SWIFT outperforms the state-of-the-art search-based method while reducing marginal per-task optimization cost by three orders of magnitude. It further generalizes to four additional unseen benchmarks and transfers successfully from GPT-4o-mini to three additional foundation models (Grok, Qwen, Gemma). Controlled ablations reveal that workflow demonstrations primarily transfer topological structure rather than surface semantics: replacing all operator names with random strings still retains over 93% of the full system's average performance. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2604_25012 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Why Search When You Can Transfer? Amortized Agentic Workflow Design from Structural Priors Du, Shiyi Liu, Jiayuan Du, Weihua Huang, Yue Li, Jiayi Luo, Yingtao Zhang, Xiangliang Conitzer, Vincent Kingsford, Carl Machine Learning Automated agentic workflow design currently relies on per-task iterative search, which is computationally prohibitive and fails to reuse structural knowledge across tasks. We observe that optimized workflows converge to a small family of domain-specific topologies, suggesting that this combinatorial search is largely redundant. Building on this insight, we propose SWIFT (Synthesizing Workflows via Few-shot Transfer), a framework that amortizes workflow design into reusable structural priors. SWIFT first distills compositional heuristics and output-interface contracts from contrastive analysis of prior search trajectories across source tasks. At inference time, it conditions a single LLM generation pass on these priors together with cross-task workflow demonstrations to synthesize a complete, executable workflow for an unseen target task, bypassing iterative search entirely. On five benchmarks, SWIFT outperforms the state-of-the-art search-based method while reducing marginal per-task optimization cost by three orders of magnitude. It further generalizes to four additional unseen benchmarks and transfers successfully from GPT-4o-mini to three additional foundation models (Grok, Qwen, Gemma). Controlled ablations reveal that workflow demonstrations primarily transfer topological structure rather than surface semantics: replacing all operator names with random strings still retains over 93% of the full system's average performance. |
| title | Why Search When You Can Transfer? Amortized Agentic Workflow Design from Structural Priors |
| topic | Machine Learning |
| url | https://arxiv.org/abs/2604.25012 |