mlsanityData to 14 Oct 2026

MetaBox-v2: A Unified Benchmark Platform for Meta-Black-Box Optimization

arXiv 2505.17745 ↗First posted 23 May 2025cs.LG cs.AI cs.NEcited by 9

Not on the map. Why?

Abstract (arXiv)

Meta-Black-Box Optimization (MetaBBO) streamlines the automation of optimization algorithm design through meta-learning. It typically employs a bi-level structure: the meta-level policy undergoes meta-training to reduce the manual effort required in developing algorithms for low-level optimization tasks. The original MetaBox (2023) provided the first open-source framework for reinforcement learning-based single-objective MetaBBO. However, its relatively narrow scope no longer keep pace with the swift advancement in this field. In this paper, we introduce MetaBox-v2 (https://github.com/MetaEvo/MetaBox) as a milestone upgrade with four novel features: 1) a unified architecture supporting RL, evolutionary, and gradient-based approaches, by which we reproduce 23 up-to-date baselines; 2) efficient parallelization schemes, which reduce the training/testing time by 10-40x; 3) a comprehensive benchmark suite of 18 synthetic/realistic tasks (1900+ instances) spanning single-objective, multi-objective, multi-model, and multi-task optimization scenarios; 4) plentiful and extensible interfaces for custom analysis/visualization and integrating to external optimization tools/benchmarks. To show the utility of MetaBox-v2, we carry out a systematic case study that evaluates the built-in baselines in terms of the optimization performance, generalization ability and learning efficiency. Valuable insights are concluded from thorough and detailed analysis for practitioners and those new to the field.

Key facts. Cited by 9 papers in the map (data through 14 Oct 2026). Builds on 42 of its 81 references that are in the map. First posted on arXiv 23 May 2025.

Builds on

42 of 81 references.

References · 42

  1. 2023Large Language Models as Optimizerscited by 864
  2. 2017Ray: A Distributed Framework for Emerging AI Applicationscited by 527
  3. 2021Tianshou: a Highly Modularized Deep Reinforcement Learning Librarycited by 143
  4. 2016COCO: A Platform for Comparing Continuous Optimizers in a Black-Box Settingcited by 138
  5. 2024LLMOPT: Learning to Define and Solve General Optimization Problems from Scratchcited by 102
  6. 2024LLaMEA: A Large Language Model Evolutionary Algorithm for Automatically Generating Metaheuristicscited by 97
  7. 2021HPO-B: A Large-Scale Reproducible Benchmark for Black-Box HPO based on OpenMLcited by 75
  8. 2024LLaMoCo: Instruction Tuning of Large Language Models for Optimization Code Generationcited by 45
  9. 2022Discovering Evolution Strategies via Meta-Black-Box Optimizationcited by 44
  10. 2019Deep Reinforcement Learning Based Parameter Control in Differential Evolutioncited by 38
  11. 2022Multi-agent Dynamic Algorithm Configurationcited by 36
  12. 2023MetaBox: A Benchmark Platform for Meta-Black-Box Optimization with Reinforcement Learningcited by 36
30 more
  1. 2023Discovering Attention-Based Genetic Algorithms via Meta-Black-Box Optimizationcited by 33
  2. 2018IOHprofiler: A Benchmarking and Profiling Tool for Iterative Optimization Heuristicscited by 28
  3. 2021Learning adaptive differential evolution algorithm from optimization experiences by policy gradientcited by 26
  4. 2024Deep Reinforcement Learning for Dynamic Algorithm Selection: A Proof-of-Principle Study on Differential Evolutioncited by 24
  5. 2024Toward Automated Algorithm Design: A Survey and Practical Guide to Meta-Black-Box-Optimizationcited by 24
  6. 2024Symbol: Generating Flexible Black-Box Optimizers through Symbolic Equation Learningcited by 22
  7. 2023Large Language Model-Enhanced Algorithm Selection: Towards Comprehensive Algorithm Representationcited by 21
  8. 2019Meta-Learning for Black-box Optimizationcited by 18
  9. 2023EvoX: A Distributed GPU-accelerated Framework for Scalable Evolutionary Computationcited by 18
  10. 2024Auto-configuring Exploration-Exploitation Tradeoff in Evolutionary Computation via Deep Reinforcement Learningcited by 17
  11. 2023B2Opt: Learning to Optimize Black-box Optimization with Little Budgetcited by 15
  12. 2024Pretrained Optimization Model for Zero-Shot Black Box Optimizationcited by 13
  13. 2025Reinforcement Learning-based Self-adaptive Differential Evolution through Automated Landscape Feature Learningcited by 12
  14. 2017Evolutionary Multitasking for Single-objective Continuous Optimization: Benchmark Problems, Performance Metric, and Baseline Resultscited by 11
  15. 2024Neural Exploratory Landscape Analysis for Meta-Black-Box-Optimizationcited by 10
  16. 2024Evolution Transformer: In-Context Evolutionary Optimizationcited by 9
  17. 2025MetaDE: Evolving Differential Evolution by Differential Evolutioncited by 9
  18. 2025Meta-Black-Box-Optimization through Offline Q-function Learningcited by 9
  19. 2022PyPop7: A Pure-Python Library for Population-Based Black-Box Optimizationcited by 7
  20. 2023MA-BBOB: Many-Affine Combinations of BBOB Functions for Evaluating AutoML Approaches in Noiseless Numerical Black-Box Optimization Contextscited by 7
  21. 2024RLEMMO: Evolutionary Multimodal Optimization Assisted By Deep Reinforcement Learningcited by 7
  22. 2024Automated Metaheuristic Algorithm Design with Autoregressive Learningcited by 7
  23. 2025Surrogate Learning in Meta-Black-Box Optimization: A Preliminary Studycited by 7
  24. 2024Using the Empirical Attainment Function for Analyzing Single-objective Black-box Optimization Algorithmscited by 6
  25. 2024Learning to Transfer for Evolutionary Multitaskingcited by 6
  26. 2025Benchmarking global optimization techniques for unmanned aerial vehicle path planningcited by 6
  27. 2025Accurate Peak Detection in Multimodal Optimization via Approximated Landscape Learningcited by 5
  28. 2020Benchmarking for Metaheuristic Black-Box Optimization: Perspectives and Open Challengescited by 4
  29. 2025Advancing CMA-ES with Learning-Based Cooperative Coevolution for Scalable Optimizationcited by 4
  30. 2025PlatMetaX: An Integrated MATLAB platform for Meta-Black-Box Optimizationcited by 1

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2025: 1; 2026: 8.

9 papers.

  1. 2026Task-free Adaptive Meta Black-box Optimizationcited by 2
  2. 2025Learning Where, What and How to Transfer: A Multi-Role Reinforcement Learning Approach for Evolutionary Multitaskingcited by 1
  3. 2026A Learning-Based Cooperative Coevolution Framework for Heterogeneous Large-Scale Global Optimizationcited by 1
  4. 2026Meta-Black-Box Optimization with Ensemble Surrogate Modeling for Robustness-Accuracy Trade-off within SAEAcited by 1
  5. 2026READY: Reward Discovery for Meta-Black-Box Optimizationcited by 0
  6. 2026SMOG: Scalable Meta-Learning for Multi-Objective Bayesian Optimizationcited by 0
  7. 2026NeurGO: Learning to Generate Elite Candidates for Meta-Black-Box Expensive Optimizationcited by 0
  8. 2026Reinforcement learning to choose optimizerscited by 0
  9. 2026Hyper Algorithm Design Agent: Evolving Learnable Optimizer from Zerocited by 0

Similar, not cited

Close in wording, no citation either way

  1. 2024ConfigX: Modular Configuration for Evolutionary Algorithms via Multitask Reinforcement Learning
  2. 2025Reinforcement learning Based Automated Design of Differential Evolution Algorithm for Black-box Optimization
  3. 2026LLMs as Adaptive Meta-Solvers: Strategy-Diverse RL for Industrial-Scale Optimization
  4. 2025DesignX: Human-Competitive Algorithm Designer for Black-Box Optimization
  5. 2024Theoretical Analysis of Meta Reinforcement Learning: Generalization Bounds and Convergence Guarantees
  6. 2025Meta-Black-Box Optimization with Bi-Space Landscape Analysis and Dual-Control Mechanism for SAEA
  7. 2024ARLBench: Flexible and Efficient Benchmarking for Hyperparameter Optimization in Reinforcement Learning
  8. 2025Scalable Multi-Objective and Meta Reinforcement Learning via Gradient Estimation