Traveling Salesperson Problem on TSP 1000 nodes
11.8Deviation from Optimal (%)GNARLBC
Evaluation results
| Method | Links | |
|---|---|---|
| GNARLBCtraining=Behavioral Cloning2025.09 | 11.8 | |
| Christofidestype=handcrafted heuristic2025.09 | 12.2 | |
| GNARLPPO+heurtraining=PPO, insertion_heuristic=true2025.09 | 20.7 | |
| Georgiev et al.beam_search_width=12802025.09 | 38.5 | |
| GNARLPPOtraining=PPO2025.09 | 45.2 |