pith. sign in

arxiv: 2607.00761 · v1 · pith:IGN5NKKEnew · submitted 2026-07-01 · 💻 cs.NE

From Consistency to Collaborative Discovery: MFEA-CoD for Multitask Novelty Search

Jiao Liu , Yanchi Li , Hua Yu , Abhishek Gupta , Yew-Soon Ong This is my paper

Pith reviewed 2026-07-02 03:23 UTC · model grok-4.3

classification 💻 cs.NE
keywords evolutionary multitaskingnovelty searchmultifactorial evolutionary algorithmcollaborative discoveryrepulsion operatoradaptive transferdeceptive optimizationmultitask optimization
0
0 comments X

The pith

MFEA-CoD coordinates multiple novelty search tasks to collaboratively discover diverse novel solutions via repulsion and adaptive transfer.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper proposes MFEA-CoD to shift evolutionary multitasking from using inter-task similarities mainly for faster convergence toward known optima to using them for joint discovery of new behaviors. It introduces a repulsion operator that pushes tasks toward different behavioral regions to cut down on repeated finds, paired with an adaptive transfer that raises or lowers sharing based on whether transferred solutions actually help the receiving task. Tests on basin landscapes, maze navigation, MuJoCo control, and generative problems show gains in diversity and better handling when objectives are deceptive. A reader would care because the method targets the common problem of search processes that either get stuck or rediscover the same things across related tasks.

Core claim

MFEA-CoD coordinates multiple novelty search tasks to collaboratively discover behaviorally novel solutions rather than merely transferring consistent search information for faster convergence. A multitask repulsion operator encourages different tasks to explore distinct regions of the unified search space, thereby reducing redundant behavioral discoveries. An adaptive inter-task transfer mechanism exploits shared discovery opportunities in overlapping novelty-improving regions by adjusting the transfer probability according to the online contribution of transferred information. MFEA-CoD is further extended to multitask novelty-augmented optimization to alleviate premature convergence caused

What carries the argument

The multitask repulsion operator paired with an adaptive inter-task transfer mechanism that together separate exploration while selectively sharing overlapping novelty gains.

If this is right

  • Reduces redundant behavioral discoveries across tasks.
  • Exploits shared opportunities in overlapping novelty regions through adjusted transfer rates.
  • Improves discovery efficiency on basin, maze, policy, and generative problems.
  • Alleviates premature convergence when novelty is added to deceptive objective functions.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

  • The same repulsion-plus-adaptive-transfer pattern could be tested on other population-based methods that currently waste evaluations on duplicate behaviors.
  • If repulsion strength is made task-dependent rather than uniform, the approach might scale to larger numbers of tasks without forcing artificial separation.
  • Applying the framework to standard objective optimization without novelty would test whether collaborative discovery ideas transfer when the goal is convergence rather than diversity.

Load-bearing premise

Overlapping novelty-improving regions exist across tasks and the repulsion operator can separate them without destroying useful shared information that the adaptive transfer would otherwise use.

What would settle it

Run the algorithm on a set of tasks whose novelty-improving regions have zero overlap and check whether performance drops below that of independent single-task novelty searches.

Figures

Figures reproduced from arXiv: 2607.00761 by Abhishek Gupta, Hua Yu, Jiao Liu, Yanchi Li, Yew-Soon Ong.

Figure 1
Figure 1. Figure 1: Illustration of multitask novelty search under three representative overlap scenarios of the target genotype sets. (a) [PITH_FULL_IMAGE:figures/full_fig_p004_1.png] view at source ↗
Figure 2
Figure 2. Figure 2: Target genotype sets of the synthetic basin-type multitask test problems [PITH_FULL_IMAGE:figures/full_fig_p009_2.png] view at source ↗
Figure 3
Figure 3. Figure 3: Illustration of the maze layouts, start positions, and target positions of the three multitask deceptive maze navigation problems. (a) [PITH_FULL_IMAGE:figures/full_fig_p010_3.png] view at source ↗
Figure 4
Figure 4. Figure 4: The considered MuJoCo benchmarks. (a) Hopper. (b) Walker2d. [PITH_FULL_IMAGE:figures/full_fig_p010_4.png] view at source ↗
Figure 5
Figure 5. Figure 5: Archive size convergence trends averaged over 20 independent runs [PITH_FULL_IMAGE:figures/full_fig_p010_5.png] view at source ↗
Figure 6
Figure 6. Figure 6: The searching distributions in the genotype space of the Novelty Search, Multitask Novelty Search, MFEA-CoD (w/o A-ITP), and MFEA-CoD on [PITH_FULL_IMAGE:figures/full_fig_p011_6.png] view at source ↗
Figure 7
Figure 7. Figure 7: The dynamics of itp1,2 and itp2,1 in MFEA-CoD on Problems 1 to 3. The values of itp1,2 and itp2,1 represent the transfer probability from Task 1 to Task 2 and from Task 2 to Task 1, respectively. induced by the proposed multitask repulsive operator, which drives the two task emitters to further expand toward their respective exclusive regions. Such a search pattern provides further evidence for the effecti… view at source ↗
Figure 8
Figure 8. Figure 8: Archive size convergence trends averaged over 20 independent [PITH_FULL_IMAGE:figures/full_fig_p012_8.png] view at source ↗
Figure 9
Figure 9. Figure 9: The final archives in the phenotype space of the Novelty Search, MAP-Elites, CMA-ME, Multitask Novelty Search, MFEA-CoD (w/o A-ITP), and [PITH_FULL_IMAGE:figures/full_fig_p013_9.png] view at source ↗
Figure 10
Figure 10. Figure 10: Best reward convergence trends averaged over 10 independent runs [PITH_FULL_IMAGE:figures/full_fig_p014_10.png] view at source ↗
Figure 11
Figure 11. Figure 11: Distribution of the archived samples and representative generated [PITH_FULL_IMAGE:figures/full_fig_p014_11.png] view at source ↗
read the original abstract

Evolutionary multitasking (EMT) has shown strong capability in solving multiple optimization problems simultaneously by exploiting latent inter-task consistency, such as similarities in promising solutions or search directions. However, most existing EMT studies remain focused on objective-driven optimization, where such consistency is mainly used to accelerate convergence toward predefined optima. In this paper, we move EMT from consistency to collaborative discovery and propose a multifactorial evolutionary algorithm with collaborative discovery (MFEA-CoD) for multitask novelty search. Unlike conventional EMT, MFEA-CoD coordinates multiple novelty search tasks to collaboratively discover behaviorally novel solutions rather than merely transferring consistent search information for faster convergence. Specifically, a multitask repulsion operator encourages different tasks to explore distinct regions of the unified search space, thereby reducing redundant behavioral discoveries. Meanwhile, an adaptive inter-task transfer mechanism exploits shared discovery opportunities in overlapping novelty-improving regions by adjusting the transfer probability according to the online contribution of transferred information. Furthermore, MFEA-CoD is extended to multitask novelty-augmented optimization, where behavioral novelty is jointly considered with objective information to alleviate premature convergence caused by deceptive objectives. Experiments on synthetic basin-type problems, deceptive maze navigation problems, MuJoCo policy optimization problems, and generative novelty search problems demonstrate that MFEA-CoD improves the efficiency of discovering diverse novel solutions and shows clear advantages in deceptive objective landscapes.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Referee Report

0 major / 3 minor

Summary. The paper proposes MFEA-CoD, a multifactorial evolutionary algorithm for multitask novelty search. It shifts EMT from objective-driven convergence to collaborative discovery by introducing a multitask repulsion operator that encourages tasks to explore distinct regions of the unified search space and an adaptive inter-task transfer mechanism that adjusts transfer probability based on the online contribution of transferred information. The method is extended to novelty-augmented optimization to mitigate premature convergence in deceptive objective landscapes. Experiments on synthetic basin-type problems, deceptive maze navigation, MuJoCo policy optimization, and generative novelty search problems are used to claim improved efficiency in discovering diverse novel solutions and advantages over standard approaches in deceptive settings.

Significance. If the empirical claims hold, the work provides a useful extension of EMT techniques into novelty search, addressing the need for diversity maintenance across tasks rather than pure convergence. The combination of repulsion for reduced redundancy and adaptive transfer for exploiting shared opportunities is a natural fit for collaborative exploration. The application to deceptive problems and the range of test domains (synthetic, maze, MuJoCo, generative) add practical relevance for evolutionary robotics and design tasks where local optima are prevalent.

minor comments (3)
  1. [Abstract] The abstract states that the adaptive mechanism 'adjusts the transfer probability according to the online contribution of transferred information,' but without the precise definition or pseudocode for this contribution metric it is difficult to assess whether the adaptation is parameter-free or introduces new hyperparameters.
  2. [Abstract] The description of the repulsion operator as encouraging 'distinct regions' would benefit from an explicit formulation (e.g., how repulsion is computed between tasks and whether it interacts with the standard EMT skill-factor assignment).
  3. The claim of 'clear advantages in deceptive objective landscapes' is presented without reference to the specific baseline algorithms or statistical tests used; adding these details would strengthen the experimental section.

Simulated Author's Rebuttal

0 responses · 0 unresolved

We thank the referee for the constructive summary, positive significance assessment, and recommendation of minor revision. No specific major comments were enumerated in the provided report, so we have no points requiring point-by-point rebuttal or clarification at this stage.

Circularity Check

0 steps flagged

No significant circularity detected

full rationale

The paper presents MFEA-CoD as an extension of standard multifactorial evolutionary algorithms by adding a multitask repulsion operator and an adaptive inter-task transfer probability. No equations or claims in the provided abstract reduce a reported performance metric or discovery result to a fitted parameter or self-citation by construction. The central method is described through explicit algorithmic components whose behavior is independent of the experimental outcomes, and the derivation chain remains self-contained against external benchmarks.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

No free parameters, axioms, or invented entities are described in the abstract; the method appears to rest on standard evolutionary operators plus two new algorithmic components whose details are not supplied.

pith-pipeline@v0.9.1-grok · 5782 in / 1079 out tokens · 49479 ms · 2026-07-02T03:23:50.587760+00:00 · methodology

discussion (0)

Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.

Reference graph

Works this paper leans on

54 extracted references · 3 canonical work pages · 3 internal anchors

  1. [1]

    Insights on transfer optimization: Because experience is the best teacher,

    A. Gupta, Y .-S. Ong, and L. Feng, “Insights on transfer optimization: Because experience is the best teacher,”IEEE Transactions on Emerging Topics in Computational Intelligence, vol. 2, no. 1, pp. 51–64, 2017

    2017

  2. [2]

    Half a dozen real-world applications of evolutionary multitasking, and more,

    A. Gupta, L. Zhou, Y .-S. Ong, Z. Chen, and Y . Hou, “Half a dozen real-world applications of evolutionary multitasking, and more,”IEEE Computational Intelligence Magazine, vol. 17, no. 2, pp. 49–66, 2022

    2022

  3. [3]

    A review on evolution- ary multitask optimization: Trends and challenges,

    T. Wei, S. Wang, J. Zhong, D. Liu, and J. Zhang, “A review on evolution- ary multitask optimization: Trends and challenges,”IEEE Transactions on Evolutionary Computation, vol. 26, no. 5, pp. 941–960, 2021

    2021

  4. [4]

    From multitask gradient descent to gradient-free evolutionary multitasking: A proof of faster convergence,

    L. Bai, W. Lin, A. Gupta, and Y .-S. Ong, “From multitask gradient descent to gradient-free evolutionary multitasking: A proof of faster convergence,”IEEE Transactions on Cybernetics, vol. 52, no. 8, pp. 8561–8573, 2021

    2021

  5. [5]

    Multifactorial evolutionary algorithm enhanced with cross-task search direction,

    J. Yin, A. Zhu, Z. Zhu, Y . Yu, and X. Ma, “Multifactorial evolutionary algorithm enhanced with cross-task search direction,” in2019 IEEE Congress on evolutionary computation (CEC). IEEE, 2019, pp. 2244– 2251

    2019

  6. [6]

    Multifactorial evolution: Toward evolutionary multitasking,

    A. Gupta, Y .-S. Ong, and L. Feng, “Multifactorial evolution: Toward evolutionary multitasking,”IEEE Transactions on Evolutionary Compu- tation, vol. 20, no. 3, pp. 343–357, 2015. JOURNAL OF LATEX CLASS FILES, VOL. 14, NO. 8, AUGUST 2015 15

    2015

  7. [7]

    Multifactorial evolu- tionary algorithm with online transfer parameter estimation: Mfea-ii,

    K. K. Bali, Y .-S. Ong, A. Gupta, and P. S. Tan, “Multifactorial evolu- tionary algorithm with online transfer parameter estimation: Mfea-ii,” IEEE Transactions on Evolutionary Computation, vol. 24, no. 1, pp. 69–83, 2019

    2019

  8. [8]

    Multiobjective multifac- torial optimization in evolutionary multitasking,

    A. Gupta, Y .-S. Ong, L. Feng, and K. C. Tan, “Multiobjective multifac- torial optimization in evolutionary multitasking,”IEEE transactions on cybernetics, vol. 47, no. 7, pp. 1652–1665, 2016

    2016

  9. [9]

    Cognizant multitasking in multiobjective multifactorial evolution: Mo-mfea-ii,

    K. K. Bali, A. Gupta, Y .-S. Ong, and P. S. Tan, “Cognizant multitasking in multiobjective multifactorial evolution: Mo-mfea-ii,”IEEE transac- tions on cybernetics, vol. 51, no. 4, pp. 1784–1796, 2020

    2020

  10. [10]

    Evolution- ary competitive multiobjective multitasking: One-pass optimization of heterogeneous pareto solutions,

    Y . Li, X. Wu, W. Gong, M. Xu, Y . Wang, and Q. Gu, “Evolution- ary competitive multiobjective multitasking: One-pass optimization of heterogeneous pareto solutions,”IEEE Transactions on Evolutionary Computation, vol. 29, no. 6, pp. 2757–2770, 2025

    2025

  11. [11]

    Explicit evolutionary multitasking for combinatorial optimization: A case study on capacitated vehicle routing problem,

    L. Feng, Y . Huang, L. Zhou, J. Zhong, A. Gupta, K. Tang, and K. C. Tan, “Explicit evolutionary multitasking for combinatorial optimization: A case study on capacitated vehicle routing problem,”IEEE transactions on cybernetics, vol. 51, no. 6, pp. 3143–3156, 2020

    2020

  12. [12]

    Solving generalized vehicle routing problem with occasional drivers via evolutionary multitasking,

    L. Feng, L. Zhou, A. Gupta, J. Zhong, Z. Zhu, K.-C. Tan, and K. Qin, “Solving generalized vehicle routing problem with occasional drivers via evolutionary multitasking,”IEEE transactions on cybernetics, vol. 51, no. 6, pp. 3171–3184, 2019

    2019

  13. [13]

    Ensemble of domain adaptation-based knowledge transfer for evolutionary multitasking,

    W. Lin, Q. Lin, L. Feng, and K. C. Tan, “Ensemble of domain adaptation-based knowledge transfer for evolutionary multitasking,” IEEE Transactions on Evolutionary Computation, vol. 28, no. 2, pp. 388–402, 2023

    2023

  14. [14]

    Evolutionary multitasking for multiobjective optimization with subspace alignment and adaptive differential evolution,

    Z. Liang, H. Dong, C. Liu, W. Liang, and Z. Zhu, “Evolutionary multitasking for multiobjective optimization with subspace alignment and adaptive differential evolution,”IEEE Transactions on Cybernetics, vol. 52, no. 4, pp. 2096–2109, 2020

    2096

  15. [15]

    Multiobjective multitask optimization with mul- tiple knowledge types and transfer adaptation,

    Y . Li and W. Gong, “Multiobjective multitask optimization with mul- tiple knowledge types and transfer adaptation,”IEEE Transactions on Evolutionary Computation, vol. 29, no. 1, pp. 205–216, 2024

    2024

  16. [16]

    Evolutionary many- task optimization based on multisource knowledge transfer,

    Z. Liang, X. Xu, L. Liu, Y . Tu, and Z. Zhu, “Evolutionary many- task optimization based on multisource knowledge transfer,”IEEE Transactions on Evolutionary Computation, vol. 26, no. 2, pp. 319–333, 2021

    2021

  17. [17]

    Knowledge structure preserving-based evolutionary many-task optimization,

    Y . Jiang, Z.-H. Zhan, K. C. Tan, S. Kwong, and J. Zhang, “Knowledge structure preserving-based evolutionary many-task optimization,”IEEE transactions on evolutionary computation, vol. 29, no. 2, pp. 287–301, 2024

    2024

  18. [18]

    Multitask evolution strategy with knowledge- guided external sampling,

    Y . Li, W. Gong, and S. Li, “Multitask evolution strategy with knowledge- guided external sampling,”IEEE Transactions on Evolutionary Compu- tation, vol. 28, no. 6, pp. 1733–1745, 2023

    2023

  19. [19]

    Jack and masters of all trades: One-pass learning sets of model sets from large pre- trained models,

    H. X. Choong, Y .-S. Ong, A. Gupta, C. Chen, and R. Lim, “Jack and masters of all trades: One-pass learning sets of model sets from large pre- trained models,”IEEE Computational Intelligence Magazine, vol. 18, no. 3, pp. 29–40, 2023

    2023

  20. [20]

    Llm2tea: An agen- tic ai designer for discovery with generative evolutionary multitasking,

    M. Wong, J. Liu, T. Rios, S. Menzel, and Y .-S. Ong, “Llm2tea: An agen- tic ai designer for discovery with generative evolutionary multitasking,” IEEE Computational Intelligence Magazine, vol. 20, no. 4, pp. 42–55, 2025

    2025

  21. [21]

    Abandoning objectives: Evolution through the search for novelty alone,

    J. Lehman and K. O. Stanley, “Abandoning objectives: Evolution through the search for novelty alone,”Evolutionary computation, vol. 19, no. 2, pp. 189–223, 2011

    2011

  22. [22]

    Discovering evolutionary stepping stones through behavior domination,

    E. Meyerson and R. Miikkulainen, “Discovering evolutionary stepping stones through behavior domination,” inProceedings of the Genetic and Evolutionary Computation Conference, 2017, pp. 139–146

    2017

  23. [23]

    Improving exploration in evolution strategies for deep reinforcement learning via a population of novelty-seeking agents,

    E. Conti, V . Madhavan, F. Petroski Such, J. Lehman, K. Stanley, and J. Clune, “Improving exploration in evolution strategies for deep reinforcement learning via a population of novelty-seeking agents,” Advances in neural information processing systems, vol. 31, 2018

    2018

  24. [24]

    Evolutionary policy optimization,

    Z. S. Mustafaoglu, K. Pingali, and R. Miikkulainen, “Evolutionary policy optimization,” inProceedings of the Genetic and Evolutionary Computation Conference Companion, 2025, pp. 735–738

    2025

  25. [25]

    Toward adaptive knowledge transfer in multifactorial evolutionary computation,

    L. Zhou, L. Feng, K. C. Tan, J. Zhong, Z. Zhu, K. Liu, and C. Chen, “Toward adaptive knowledge transfer in multifactorial evolutionary computation,”IEEE transactions on cybernetics, vol. 51, no. 5, pp. 2563–2576, 2020

    2020

  26. [26]

    Adaptive multifac- torial evolutionary optimization for multitask reinforcement learning,

    A. D. Martinez, J. Del Ser, E. Osaba, and F. Herrera, “Adaptive multifac- torial evolutionary optimization for multitask reinforcement learning,” IEEE Transactions on Evolutionary Computation, vol. 26, no. 2, pp. 233–247, 2021

    2021

  27. [27]

    Multiobjective multitask optimization with mul- tiple knowledge types and transfer adaptation,

    Y . Li and W. Gong, “Multiobjective multitask optimization with mul- tiple knowledge types and transfer adaptation,”IEEE Transactions on Evolutionary Computation, vol. 29, no. 1, pp. 205–216, 2025

    2025

  28. [28]

    Affine transformation-enhanced multifactorial optimization for heterogeneous problems,

    X. Xue, K. Zhang, K. C. Tan, L. Feng, J. Wang, G. Chen, X. Zhao, L. Zhang, and J. Yao, “Affine transformation-enhanced multifactorial optimization for heterogeneous problems,”IEEE Transactions on Cy- bernetics, vol. 52, no. 7, pp. 6217–6231, 2020

    2020

  29. [29]

    Optimal transport-based distributional pairing in transfer multiobjective optimization,

    J. Liu, W. Liu, J. T. W. En, C. Chen, P. S. Tan, and Y .-S. Ong, “Optimal transport-based distributional pairing in transfer multiobjective optimization,”IEEE Transactions on Evolutionary Computation, 2025

    2025

  30. [30]

    Multiobjective multitask optimization with manifold structure-driven knowledge transfer,

    T. Zhang, X. Wu, Y . Li, S. Li, and W. Gong, “Multiobjective multitask optimization with manifold structure-driven knowledge transfer,”IEEE Transactions on Systems, Man, and Cybernetics: Systems, 2026

    2026

  31. [31]

    A surrogate-assisted evolutionary framework for expensive multitask optimization problems,

    S. Tan, Y . Wang, G. Sun, T. Pang, and K. Tang, “A surrogate-assisted evolutionary framework for expensive multitask optimization problems,” IEEE Transactions on Evolutionary Computation, 2024

    2024

  32. [32]

    Multiproblem surrogates: Transfer evolutionary multiobjective optimization of com- putationally expensive problems,

    A. T. W. Min, Y .-S. Ong, A. Gupta, and C.-K. Goh, “Multiproblem surrogates: Transfer evolutionary multiobjective optimization of com- putationally expensive problems,”IEEE Transactions on Evolutionary Computation, vol. 23, no. 1, pp. 15–28, 2017

    2017

  33. [33]

    Extremo: Transfer evolutionary multiobjective optimization with proof of faster convergence,

    J. Liu, A. Gupta, C. Ooi, and Y .-S. Ong, “Extremo: Transfer evolutionary multiobjective optimization with proof of faster convergence,”IEEE Transactions on Evolutionary Computation, vol. 29, no. 1, pp. 102–116, 2024

    2024

  34. [34]

    Transfer stacking from low- to high-fidelity: A surrogate-assisted bi-fidelity evolutionary algorithm,

    H. Wang, Y . Jin, C. Yang, and L. Jiao, “Transfer stacking from low- to high-fidelity: A surrogate-assisted bi-fidelity evolutionary algorithm,” Applied soft computing, vol. 92, p. 106276, 2020

    2020

  35. [35]

    (θ l,θ u)-parametric multi-task optimization: Joint search in solution and infinite task spaces,

    T. Wei, J. Liu, A. Gupta, P. S. Tan, and Y .-S. Ong, “(θ l,θ u)-parametric multi-task optimization: Joint search in solution and infinite task spaces,” IEEE Transactions on Evolutionary Computation, 2025

    2025

  36. [36]

    Parametric pareto set learning for expensive multi-objective optimization,

    J. Cheng, B. Xue, and Q. Zhang, “Parametric pareto set learning for expensive multi-objective optimization,” inProceedings of the AAAI Conference on Artificial Intelligence, vol. 40, no. 43, 2026, pp. 36 829– 36 837

    2026

  37. [37]

    Amortized Multi-Objective Optimization Across Tasks with Generative Solution Modeling

    T. Wei, J. Liu, A. Gupta, C. C. Ooi, P. S. Tan, and Y .-S. Ong, “Parametric expensive multi-objective optimization via generative solution model- ing,”arXiv preprint arXiv:2511.09598, 2025

    work page internal anchor Pith review Pith/arXiv arXiv 2025

  38. [38]

    Bayesian inverse transfer in evolu- tionary multiobjective optimization,

    J. Liu, A. Gupta, and Y .-S. Ong, “Bayesian inverse transfer in evolu- tionary multiobjective optimization,”ACM Transactions on Evolutionary Learning, vol. 4, no. 4, pp. 1–27, 2025

    2025

  39. [39]

    Bayesian forward- inverse transfer for multiobjective optimization,

    T. Wei, J. Liu, A. Gupta, P. S. Tan, and Y .-S. Ong, “Bayesian forward- inverse transfer for multiobjective optimization,” inInternational Con- ference on Parallel Problem Solving from Nature. Springer, 2024, pp. 135–152

    2024

  40. [40]

    Critical factors in the performance of novelty search,

    S. Kistemaker and S. Whiteson, “Critical factors in the performance of novelty search,” inProceedings of the 13th annual conference on Genetic and evolutionary computation, 2011, pp. 965–972

    2011

  41. [41]

    Covari- ance matrix adaptation for the rapid illumination of behavior space,

    M. C. Fontaine, J. Togelius, S. Nikolaidis, and A. K. Hoover, “Covari- ance matrix adaptation for the rapid illumination of behavior space,” inProceedings of the 2020 genetic and evolutionary computation conference, 2020, pp. 94–102

    2020

  42. [42]

    Evolving a diversity of virtual creatures through novelty search and local competition,

    J. Lehman and K. O. Stanley, “Evolving a diversity of virtual creatures through novelty search and local competition,” inProceedings of the 13th annual conference on Genetic and evolutionary computation, 2011, pp. 211–218

    2011

  43. [43]

    Simulated binary crossover for continuous search space,

    K. Deb, R. B. Agrawalet al., “Simulated binary crossover for continuous search space,”Complex systems, vol. 9, no. 2, pp. 115–148, 1995

    1995

  44. [44]

    pyribs: A bare-bones python library for quality diversity optimization,

    B. Tjanaka, M. C. Fontaine, D. H. Lee, Y . Zhang, N. R. Balam, N. Dennler, S. S. Garlanka, N. D. Klapsis, and S. Nikolaidis, “pyribs: A bare-bones python library for quality diversity optimization,” in Proceedings of the Genetic and Evolutionary Computation Conference, 2023, pp. 220–229

    2023

  45. [45]

    Kheperax: a lightweight jax-based robot control environment for benchmarking quality-diversity algorithms,

    L. Grillotti and A. Cully, “Kheperax: a lightweight jax-based robot control environment for benchmarking quality-diversity algorithms,” in Proceedings of the Companion Conference on Genetic and Evolutionary Computation, 2023, pp. 2163–2165

    2023

  46. [46]

    Mujoco: A physics engine for model- based control,

    E. Todorov, T. Erez, and Y . Tassa, “Mujoco: A physics engine for model- based control,” in2012 IEEE/RSJ international conference on intelligent robots and systems. IEEE, 2012, pp. 5026–5033

    2012

  47. [47]

    Analyzing and improving the image quality of stylegan,

    T. Karras, S. Laine, M. Aittala, J. Hellsten, J. Lehtinen, and T. Aila, “Analyzing and improving the image quality of stylegan,” inProceedings of the IEEE/CVF conference on computer vision and pattern recognition, 2020, pp. 8110–8119

    2020

  48. [48]

    Exploring deceptive mazes with novelty search,

    B. Tjanaka, “Exploring deceptive mazes with novelty search,” pyribs.org, 2025. [Online]. Available: https://docs.pyribs.org/en/stable/ tutorials/ns maze.html

    2025

  49. [49]

    Gymnasium: A Standard Interface for Reinforcement Learning Environments

    M. Towers, A. Kwiatkowski, J. Terry, J. U. Balis, G. De Cola, T. Deleu, M. Goul ˜ao, A. Kallinteris, M. Krimmel, A. KGet al., “Gymnasium: A standard interface for reinforcement learning environments,”arXiv preprint arXiv:2407.17032, 2024. JOURNAL OF LATEX CLASS FILES, VOL. 14, NO. 8, AUGUST 2015 16

    work page internal anchor Pith review Pith/arXiv arXiv 2024

  50. [50]

    Learning transferable visual models from natural language supervision,

    A. Radford, J. W. Kim, C. Hallacy, A. Ramesh, G. Goh, S. Agarwal, G. Sastry, A. Askell, P. Mishkin, J. Clarket al., “Learning transferable visual models from natural language supervision,” inInternational conference on machine learning. PmLR, 2021, pp. 8748–8763

    2021

  51. [51]

    Kernel density estimation and its application,

    S. Weglarczyk, “Kernel density estimation and its application,” inITM web of conferences, vol. 23. EDP Sciences, 2018, p. 00037

    2018

  52. [52]

    Illuminating search spaces by mapping elites

    J.-B. Mouret and J. Clune, “Illuminating search spaces by mapping elites,”arXiv preprint arXiv:1504.04909, 2015

    work page internal anchor Pith review Pith/arXiv arXiv 2015

  53. [53]

    Reducing the time complexity of the derandomized evolution strategy with covariance matrix adaptation (cma-es),

    N. Hansen, S. D. M ¨uller, and P. Koumoutsakos, “Reducing the time complexity of the derandomized evolution strategy with covariance matrix adaptation (cma-es),”Evolutionary computation, vol. 11, no. 1, pp. 1–18, 2003. JOURNAL OF LATEX CLASS FILES, VOL. 14, NO. 8, AUGUST 2015 17 Supplementary File of “From Consistency to Collaborative Discovery:MFEA-CoDfo...

    2003

  54. [54]

    The MFEA-CoD settings use novelty weightλ= 0.3, linear decayed repulsion strength from 0.5 to 0, regularization coefficient 0.5, adaptive-ITPlearning rate 0.1, and the initial transfer probability itp 1,2 =itp 2,1 = 0.3