Traveling Salesman Problem on TSPLib <100
1.07Optimality Gap (Before)INViT-1V+ASAP w/ MAML
Evaluation results
| Method | Links | ||||
|---|---|---|---|---|---|
| INViT-1V+ASAP w/ MAMLBackbone=INViT-1V, Adaptation=ASAP, Algorithm=MAML2025.01 | 1.07 | — | 0.99 | 0.08 | |
| POMO+ASAP w/ MAMLBackbone=POMO, Adaptation=ASAP, Algorithm=MAML2025.01 | 1.11 | — | 0.99 | 0.12 | |
| POMO+ASAP w/o MAMLBackbone=POMO, Adaptation=ASAP, Algorithm=None2025.01 | 1.28 | — | 1.03 | 0.25 | |
| INViT-1V+ASAP w/o MAMLBackbone=INViT-1V, Adaptation=ASAP, Algorithm=None2025.01 | 1.36 | — | 1.07 | 0.29 | |
| POMO+MAMLBackbone=POMO, Adaptation=MAML2025.01 | 1.47 | — | 1.32 | 0.15 | |
| INViT-1V+MAMLBackbone=INViT-1V, Adaptation=MAML2025.01 | 1.77 | — | 1.32 | 0.45 | |
| POMOBackbone=POMO, Adaptation=None2025.01 | 1.92 | — | 1.85 | 0.07 | |
| INViT-1VBackbone=INViT-1V, Adaptation=None2025.01 | 2.65 | — | 2.41 | 0.24 | |
| LEHDDecoding Strategy=greedy2026.05 | — | 1.064 | — | — | |
| MnLPDecoding Strategy=greedy2026.05 | — | 0.673 | — | — | |
| POMODecoding Strategy=aug x 82026.05 | — | 0.792 | — | — |