Unify ML4TSP: Drawing Methodological Principles for TSP and Beyond from Streamlined Design Space of Learning and Search

Yang Li, Jiale Ma, Wenzheng Pan, Runzhong Wang, Haoyu Geng +2 more
2/12/2026

No official code found. Best implementation: Thinklab-SJTU/ML4TSPBench (45 stars, last commit 2025-05-20).

Semantic Scholar

Code Implementations(1)

Thinklab-SJTU/ML4TSPBenchCommunity73%

Official implementation of ICLR 2025 paper: "Unify ML4TSP: Drawing Methodological Principles for TSP and Beyond from Streamlined Design Space of Learning and Search".

450C, C++Jan 22, 20251 years ago
combinatorial-optimizationdiffusion-modelsmachine-learningmcts-algorithmtsp-problem

Cite this paper

@article{li2026unify,
  title  = {Unify ML4TSP: Drawing Methodological Principles for TSP and Beyond from Streamlined Design Space of Learning and Search},
  author = {Yang Li and Jiale Ma and Wenzheng Pan and Runzhong Wang and Haoyu Geng and Nianzu Yang and Junchi Yan},
  year   = {2026},
  url    = {https://api.semanticscholar.org/CorpusID:6b24492dbf3e5c80ae6b080edde72e69af654e89},
  journal = {ICLR 2025 2025}
}

Discussion