HMACE: Heterogeneous Multi-Agent Collaborative Evolution for Combinatorial Optimization
Yuping Yan, Jirui Han, Fei Ming, Yuanshuai Li, Yaochu Jin · 2026
Read HMACE: Heterogeneous Multi-Agent Collaborative Evolution for Combinatorial Optimization by Yuping Yan, Jirui Han, Fei Ming, Yuanshuai Li, Yaochu Jin (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2605-07214.txt.
Description
Yuping Yan and four co-authors let a team of language-model agents design solution methods for hard optimization problems. Earlier model-based heuristic design used rigid single workflows that converged too early and explored poorly. HMACE instead splits each generation among four specialized agents that propose, generate, evaluate, and reflect.
Behavior-aware retrieval, light filtering, and a fitness-based archive keep the search diverse and avoid wasted evaluations. Across traveling salesman, packing, and scheduling problems, the framework achieved a strong quality-to-cost balance against single-agent and multi-agent rivals. On two tasks it posted the lowest average gaps while spending far fewer tokens than baselines.
Catalog record
| CREATOR | Yuping Yan, Jirui Han, Fei Ming, Yuanshuai Li, Yaochu Jin |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 63 KB |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2605.07214