Saved in:
| Main Authors: | , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2603.22791 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866911541012463616 |
|---|---|
| author | Song, Weijia Yue, Jiashu Pang, Zhe |
| author_facet | Song, Weijia Yue, Jiashu Pang, Zhe |
| contents | How should multi-agent systems be designed, and can that design knowledge be captured in a form that is inspectable, revisable, and transferable? We introduce ABSTRAL, a framework that treats MAS architecture as an evolving natural-language document, an artifact refined through contrastive trace analysis. Three findings emerge. First, we provide a precise measurement of the multi-agent coordination tax: under fixed turn budgets, ensembles achieve only 26% turn efficiency, with 66% of tasks exhausting the limit, yet still improve over single-agent baselines by discovering parallelizable task decompositions. Second, design knowledge encoded in documents transfers: topology reasoning and role templates learned on one domain provide a head start on new domains, with transferred seeds matching coldstart iteration 3 performance in a single iteration. Third, contrastive trace analysis discovers specialist roles absent from any initial design, a capability no prior system demonstrates. On SOPBench (134 bank tasks, deterministic oracle), ABSTRAL reaches 70% validation / 65.96% test pass rate with a GPT-4o backbone. We release the converged documents as inspectable design rationale. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2603_22791 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | ABSTRAL: Automatic Design of Multi-Agent Systems Through Iterative Refinement and Topology Optimization Song, Weijia Yue, Jiashu Pang, Zhe Artificial Intelligence How should multi-agent systems be designed, and can that design knowledge be captured in a form that is inspectable, revisable, and transferable? We introduce ABSTRAL, a framework that treats MAS architecture as an evolving natural-language document, an artifact refined through contrastive trace analysis. Three findings emerge. First, we provide a precise measurement of the multi-agent coordination tax: under fixed turn budgets, ensembles achieve only 26% turn efficiency, with 66% of tasks exhausting the limit, yet still improve over single-agent baselines by discovering parallelizable task decompositions. Second, design knowledge encoded in documents transfers: topology reasoning and role templates learned on one domain provide a head start on new domains, with transferred seeds matching coldstart iteration 3 performance in a single iteration. Third, contrastive trace analysis discovers specialist roles absent from any initial design, a capability no prior system demonstrates. On SOPBench (134 bank tasks, deterministic oracle), ABSTRAL reaches 70% validation / 65.96% test pass rate with a GPT-4o backbone. We release the converged documents as inspectable design rationale. |
| title | ABSTRAL: Automatic Design of Multi-Agent Systems Through Iterative Refinement and Topology Optimization |
| topic | Artificial Intelligence |
| url | https://arxiv.org/abs/2603.22791 |