SPO: Discovering Adaptive Large Neighborhood Search Operators via Stackelberg Program Optimization
This work introduces Stackelberg Program Optimization (SPO), an LLM-based framework for discovering adaptive executable destroy-repair programs that outperforms strong baselines across a broad range of settings and generalizes beyond the discovery scale to larger instances and benchmark sets.
Xin-Yi Ke, Kai Li, Jun-Liang Xing et al.
· 0 citations