Rethinking Efficiency in Neural Combinatorial Optimization: Batched Preference Optimization with Mamba
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , , , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2026
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866910171599470592 |
|---|---|
| author | Xu, Zhenxing Ma, Zeyuan Bao, Weidong Zheng, Yan Wang, Ji Cao, Zhiguang |
| author_facet | Xu, Zhenxing Ma, Zeyuan Bao, Weidong Zheng, Yan Wang, Ji Cao, Zhiguang |
| contents | We study efficiency as a first-class objective in Neural Combinatorial Optimization (NCO) and present ECO, an efficient learning framework that combines batched preference optimization with a Mamba backbone. Instead of tightly interleaving every policy update with on-policy rollouts, ECO decouples trajectory generation from gradient updates through two stages: supervised warm-up on pre-computed solutions and iterative Direct Preference Optimization (DPO) on batched candidate sets generated by the current policy. We pair this learning pipeline with a mixed Mamba encoder-decoder that reduces memory growth on long sequences and improves hardware utilization. A local-search-guided bootstrapping strategy is further used during training to widen preference margins and stabilize iterative improvement. Importantly, local search is only used to construct stronger preference pairs during training and is never invoked at inference time. On TSP and CVRP, ECO achieves the strongest overall performance among the compared neural baselines while also delivering clear advantages in memory usage and throughput. We provide additional analysis on memory scaling, throughput, and the contribution of each design component. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2602_20730 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Rethinking Efficiency in Neural Combinatorial Optimization: Batched Preference Optimization with Mamba Xu, Zhenxing Ma, Zeyuan Bao, Weidong Zheng, Yan Wang, Ji Cao, Zhiguang Machine Learning We study efficiency as a first-class objective in Neural Combinatorial Optimization (NCO) and present ECO, an efficient learning framework that combines batched preference optimization with a Mamba backbone. Instead of tightly interleaving every policy update with on-policy rollouts, ECO decouples trajectory generation from gradient updates through two stages: supervised warm-up on pre-computed solutions and iterative Direct Preference Optimization (DPO) on batched candidate sets generated by the current policy. We pair this learning pipeline with a mixed Mamba encoder-decoder that reduces memory growth on long sequences and improves hardware utilization. A local-search-guided bootstrapping strategy is further used during training to widen preference margins and stabilize iterative improvement. Importantly, local search is only used to construct stronger preference pairs during training and is never invoked at inference time. On TSP and CVRP, ECO achieves the strongest overall performance among the compared neural baselines while also delivering clear advantages in memory usage and throughput. We provide additional analysis on memory scaling, throughput, and the contribution of each design component. |
| title | Rethinking Efficiency in Neural Combinatorial Optimization: Batched Preference Optimization with Mamba |
| topic | Machine Learning |
| url | https://arxiv.org/abs/2602.20730 |