Research map / Label-efficient and robust learning / Test-time adaptation
Meta-SAGE: Scale Meta-Learning Scheduled Adaptation with Guided Exploration for Mitigating Scale Shift on Combinatorial Optimization
Jiwoo Son, Minsu Kim, Hyeonah Kim, Jinkyoo Park
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 Meta-SAGE instead of EAS | POMO · CVRP, N=1,000 | Gap: 3.95% -> 0.683% | Improved |
| Use Meta-SAGE instead of EAS | Sym-NCO · Real-world CVRP, 500 ≤ N < 600 | Objective: 95,384 -> 94,028 | Improved |
| Add scale meta-learning (SML) (ablation) | Sym-NCO with test-time adaptation · TSP, N=500 | Objective: 18.63 -> 17.70 | Improved |
| Add locality bias (ablation) | Sym-NCO with SML · TSP, N=500 | Objective: 17.70 -> 17.68 | Improved |
| Schedule decay of locality bias (ablation) | Sym-NCO with SML and locality bias · TSP, N=500 | Objective: 17.68 -> 17.67 | Improved |
| Schedule SoftMax temperature (ablation) | Sym-NCO with SML and scheduled locality bias · TSP, N=500 | Objective: 17.67 -> 17.22 | Improved |
| Add SML to transform the initial embedding (ablation) | SAGE · TSP, N=500 | L1 embedding distance: 8,066.13 -> 7,175.99 | Improved |
| Add SML to transform the initial embedding (ablation) | SAGE · CVRP, N=500 | L1 embedding distance: 10,361.51 -> 9,285.22 | Improved |
Closest papers by meaning
- MetaBox-v2: A Unified Benchmark Platform for Meta-Black-Box Optimization (NeurIPS 2025 · similarity 0.87)
- Meta-Reinforcement Learning with Universal Policy Adaptation: Provable Near-Optimality under All-task Optimum Comparator (NeurIPS 2024 · similarity 0.86)
- Meta-Reinforcement Learning via Exploratory Task Clustering (AAAI 2024 · similarity 0.86)
- AMAGO-2: Breaking the Multi-Task Barrier in Meta-Reinforcement Learning with Transformers (NeurIPS 2024 · similarity 0.86)
- MetaCARD: Meta-Reinforcement Learning with Task Uncertainty Feedback via Decoupled Context-Aware Reward and Dynamics Components (AAAI 2024 · similarity 0.86)
- Can Learned Optimization Make Reinforcement Learning Less Difficult? (NeurIPS 2024 · similarity 0.85)
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.
