Research map / Graph learning / GNN architectures
Beyond Simple Graphs: Neural Multi-Objective Routing on Multigraphs
Filip Rydin, Attila Lischka, Jiaming Wu, Morteza Haghir Chehreghani, Balazs Kulcsar
Read the paperCheck your idea against 137,217 papers
AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.
Measured effects reported in the paper
| What was changed | Where | Measured | Effect |
|---|---|---|---|
| use GMS-EB instead of MBM | MBM · MOCVRP XASY100 | normalized hypervolume: 0.60 -> 0.79 | Improved |
| use GMS-DH instead of MBM | MBM · MGMOCVRP FLEX2-100 | normalized hypervolume: 0.81 -> 0.85 | Improved |
| use GMS-EB instead of MBM | MBM · MGMOTSPTW FIX2-50 | normalized hypervolume: 0.64 -> 0.93 | Improved |
| add instance augmentation to GMS-EB | GMS-EB · MOTSP TMAT50 | HV gap: 1.50% -> 1.14% | Improved |
| add instance augmentation to GMS-EB | GMS-EB · MOTSP TMAT100 | inference time: 3.6m -> 29m | Hurt |
| add pre-pruning before encoder to GMS-DH (ablation) | GMS-DH · MGMOTSP FLEX2-50 | normalized hypervolume: 0.89 -> 0.77 | Hurt |
| add pre-pruning before encoder to GMS-DH (ablation) | GMS-DH · MGMOCVRP FLEX2-50 | normalized hypervolume: 0.85 -> 0.75 | Hurt |
| add learned selection-decoder instead of simple linear pruning (ablation) | GMS-DH Simple · MGMOTSPTW FLEX2-50 | normalized hypervolume: 0.83 -> 0.85 | Improved |
| add learned selection-decoder instead of simple linear pruning (ablation) | GMS-DH Simple · MGMOTSPTW FIX2-50 | normalized hypervolume: 0.85 -> 0.91 | Improved |
| use GMS-DH instead of GMS-EB | GMS-EB · MGMOTSP FLEX2-100 | inference time: 9.1m -> 49s | Improved |
| use LKH instead of GMS-EB | GMS-EB · MOTSP TMAT50 | HV gap: 1.50% -> 0.00% | Mixed |
| use GMS-DH instead of GMS-EB | GMS-EB · MOTSP TMAT200 (zero-shot) | normalized hypervolume: 0.66 -> 0.65 | Mixed |
Closest papers by meaning
- Neural Link Prediction with Walk Pooling (ICLR 2022 · similarity 0.87)
- Neural Execution of Graph Algorithms (ICLR 2020 · similarity 0.86)
- Semantic Routing via Autoregressive Modeling (NeurIPS 2024 · similarity 0.86)
- Improving Graph Neural Networks by Learning Continuous Edge Directions (ICLR 2025 · similarity 0.85)
- Learning Long Range Dependencies on Graphs via Random Walks (ICLR 2025 · similarity 0.85)
- GraphTOP: Graph Topology-Oriented Prompting for Graph Neural Networks (NeurIPS 2025 · similarity 0.84)
Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.
