mlsanityData to 14 Oct 2026

Generating and Blending Game Levels via Quality-Diversity in the Latent Space of a Variational Autoencoder

arXiv 2102.12463 ↗First posted 24 Feb 2021cs.LG cs.AIcited by 27

Not on the map. Why?

Abstract (arXiv)

Several works have demonstrated the use of variational autoencoders (VAEs) for generating levels in the style of existing games and blending levels across different games. Further, quality-diversity (QD) algorithms have also become popular for generating varied game content by using evolution to explore a search space while focusing on both variety and quality. To reap the benefits of both these approaches, we present a level generation and game blending approach that combines the use of VAEs and QD algorithms. Specifically, we train VAEs on game levels and run the MAP-Elites QD algorithm using the learned latent space of the VAE as the search space. The latent space captures the properties of the games whose levels we want to generate and blend, while MAP-Elites searches this latent space to find a diverse set of levels optimizing a given objective such as playability. We test our method using models for 5 different platformer games as well as a blended domain spanning 3 of these games. We refer to using MAP-Elites for blending as Blend-Elites. Our results show that MAP-Elites in conjunction with VAEs enables the generation of a diverse set of playable levels not just for each individual game but also for the blended domain while illuminating game-specific regions of the blended latent space.

Key facts. Cited by 27 papers in the map (data through 14 Oct 2026). Builds on 30 of its 56 references that are in the map. First posted on arXiv 24 Feb 2021.

Builds on

30 of 56 references.

References · 30

  1. 2013Auto-Encoding Variational Bayescited by 4,053
  2. 2015Illuminating search spaces by mapping elitescited by 425
  3. 2019Go-Explore: a New Approach for Hard-Exploration Problemscited by 212
  4. 2017Procedural Content Generation via Machine Learning (PCGML)cited by 137
  5. 2019Covariance Matrix Adaptation for the Rapid Illumination of Behavior Spacecited by 78
  6. 2018Evolving Mario Levels in the Latent Space of a Deep Convolutional Generative Adversarial Networkcited by 76
  7. 2020Illuminating Mario Scenes in the Latent Space of a Generative Adversarial Networkcited by 52
  8. 2019Procedural Content Generation through Quality Diversitycited by 45
  9. 2016The VGLC: The Video Game Level Corpuscited by 41
  10. 2016Super Mario as a String: Platformer Level Generation Via LSTMscited by 40
  11. 2019Mapping Hearthstone Deck Spaces through MAP-Elites with Sliding Boundariescited by 32
  12. 2018Talakat: Bullet Hell Generation through Constrained Map-Elitescited by 26
18 more
  1. 2020Interactive Evolution and Exploration Within Latent Level-Design Space of Generative Adversarial Networkscited by 22
  2. 2020Baba is Y'all: Collaborative Mixed-Initiative Level Designcited by 20
  3. 2020CPPN2GAN: Combining Compositional Pattern Producing Networks and GANs for Large-scale Pattern Generationcited by 19
  4. 2019Empowering Quality Diversity in Dungeon Design with Interactive Constrained MAP-Elitescited by 18
  5. 2019Diverse Agents for Ad-Hoc Cooperation in Hanabicited by 17
  6. 2020Conditional Level Generation and Game Blendingcited by 17
  7. 2018Automated Game Design via Conceptual Expansioncited by 15
  8. 2020Controllable Level Blending between Games using Variational Autoencoderscited by 14
  9. 2020Finding Game Levels with the Right Difficulty in a Few Trials through Intelligent Trial-and-Errorcited by 14
  10. 2020Sequential Segment-based Level Generation and Blending using Variational Autoencoderscited by 14
  11. 2020Exploring Level Blending across Platformers via Paths and Affordancescited by 14
  12. 2020Mech-Elites: Illuminating the Mechanic Space of GVGAIcited by 13
  13. 2020Towards Game Design via Creative Machine Learning (GDCML)cited by 13
  14. 2020Multi-Domain Level Generation and Blending with Sketches via Example-Driven BSP and Variational Autoencoderscited by 11
  15. 2018Generating Levels That Teach Mechanicscited by 9
  16. 2020Game Level Clustering and Generation using Gaussian Mixture VAEscited by 9
  17. 2020Entity Embedding as Game Representationcited by 8
  18. 2017Mechanics Automatically Recognized via Interactive Observation: Jumpingcited by 4

Built on by

Cited by year (2026 to 14 Oct)

Citing papers per year: 2021: 3; 2022: 7; 2023: 9; 2024: 2; 2025: 4; 2026: 2.

27 papers.

Sort:

  1. 2021Differentiable Quality Diversitycited by 72
  2. 2023MarioGPT: Open-Ended Text2Level Generation through Large Language Modelscited by 64
  3. 2022Deep Surrogate Assisted Generation of Environmentscited by 40
  4. 2022Covariance Matrix Adaptation MAP-Annealingcited by 25
  5. 2023Proximal Policy Gradient Arborescence for Quality Diversity Reinforcement Learningcited by 20
  6. 2023The Five-Dollar Model: Generating Game Maps and Sprites from Sentence Embeddingscited by 14
  7. 2021Dungeon and Platformer Level Blending and Generation using Conditional VAEscited by 10
  8. 2023Surrogate Assisted Generation of Human-Robot Interaction Scenarioscited by 8
  9. 2022Online Damage Recovery for Physical Robots with Hierarchical Quality-Diversitycited by 7
  10. 2023Lode Encoder: AI-constrained co-creativitycited by 6
  11. 2022Hierarchical Quality-Diversity for Online Damage Recoverycited by 5
  12. 2022Mario Plays on a Manifold: Generating Functional Content in Latent Space through Differential Geometrycited by 5
15 more
  1. 2023Evolving Populations of Diverse RL Agents with MAP-Elitescited by 5
  2. 2024Measuring Diversity of Game Scenarioscited by 5
  3. 2025Cooking Up Creativity: Enhancing LLM Creativity through Structured Recombinationcited by 3
  4. 2021Ensemble Learning For Mega Man Level Generationcited by 1
  5. 2022Clustering-based Tile Embedding (CTE): A General Representation for Level Design with Skewed Tile Distributionscited by 1
  6. 2023Joint Level Generation and Translation Using Gameplay Videoscited by 1
  7. 2023Reconstructing Existing Levels through Level Inpaintingcited by 1
  8. 2025Level Generation with Constrained Expressive Rangecited by 1
  9. 2025Video Game Level Design as a Multi-Agent Reinforcement Learning Problemcited by 1
  10. 2026Discount Model Search for Quality Diversity Optimization in High-Dimensional Measure Spacescited by 1
  11. 2026Multiverse: Language-Conditioned Multi-Game Level Blending via Shared Representationcited by 1
  12. 2022Latent Combinational Game Designcited by 0
  13. 2023Targeted Analysis of High-Risk States Using an Oriented Variational Autoencodercited by 0
  14. 2024LVNS-RAVE: Diversified audio generation with RAVE and Latent Vector Novelty Searchcited by 0
  15. 2025Level Generation with Quantum Reservoir Computingcited by 0