mlsanityData to 7 Oct 2026

CausalVerse: Benchmarking Causal Representation Learning with Configurable High-Fidelity Simulations

arXiv 2510.14049 ↗First posted 15 Oct 2025cs.LG cs.MScited by 3

Not on the map. Why?

Abstract (arXiv)

Causal Representation Learning (CRL) aims to uncover the data-generating process and identify the underlying causal variables and relations, whose evaluation remains inherently challenging due to the requirement of known ground-truth causal variables and causal structure. Existing evaluations often rely on either simplistic synthetic datasets or downstream performance on real-world tasks, generally suffering a dilemma between realism and evaluative precision. In this paper, we introduce a new benchmark for CRL using high-fidelity simulated visual data that retains both realistic visual complexity and, more importantly, access to ground-truth causal generating processes. The dataset comprises around 200 thousand images and 3 million video frames across 24 sub-scenes in four domains: static image generation, dynamic physical simulations, robotic manipulations, and traffic situation analysis. These scenarios range from static to dynamic settings, simple to complex structures, and single to multi-agent interactions, offering a comprehensive testbed that hopefully bridges the gap between rigorous evaluation and real-world applicability. In addition, we provide flexible access to the underlying causal structures, allowing users to modify or configure them to align with the required assumptions in CRL, such as available domain labels, temporal dependencies, or intervention histories. Leveraging this benchmark, we evaluated representative CRL methods across diverse paradigms and offered empirical insights to assist practitioners and newcomers in choosing or extending appropriate CRL frameworks to properly address specific types of real problems that can benefit from the CRL perspective. Welcome to visit our: Project page:https://causal-verse.github.io/, Dataset:https://huggingface.co/CausalVerse.

Key facts. Cited by 3 papers in the map (data through 7 Oct 2026). Builds on 33 of its 66 references that are in the map. First posted on arXiv 15 Oct 2025.

Builds on

33 of 66 references.

References · 33

  1. 2021High-Resolution Image Synthesis with Latent Diffusion Modelscited by 18,694
  2. 2017CARLA: An Open Urban Driving Simulatorcited by 2,431
  3. 2024CogVideoX: Text-to-Video Diffusion Models with An Expert Transformercited by 2,198
  4. 2019The Replica Dataset: A Digital Replica of Indoor Spacescited by 885
  5. 2019Variational Autoencoders and Nonlinear ICA: A Unifying Frameworkcited by 494
  6. 2020robosuite: A Modular Simulation Framework and Benchmark for Robot Learningcited by 484
  7. 2021Self-Supervised Learning with Data Augmentations Provably Isolates Content from Stylecited by 308
  8. 2016Unsupervised Feature Extraction by Time-Contrastive Learning and Nonlinear ICAcited by 270
  9. 2021Contrastive Learning Inverts the Data Generating Processcited by 220
  10. 2021Disentanglement via Mechanism Sparsity Regularization: A New Principle for Nonlinear ICAcited by 158
  11. 2022Weakly supervised causal representation learningcited by 140
  12. 2021Learning Temporally Causal Latent Processes from General Temporal Datacited by 96
21 more
  1. 2023Learning Linear Causal Representations from Interventions under General Nonlinear Mixingcited by 92
  2. 2022On the Identifiability of Nonlinear ICA: Sparsity and Beyondcited by 84
  3. 2020Causal Discovery in Physical Systems from Videoscited by 77
  4. 2023Multi-View Causal Representation Learning with Partial Observabilitycited by 72
  5. 2022Temporally Disentangled Representation Learningcited by 59
  6. 2021Learning latent causal graphs via mixture oraclescited by 55
  7. 2022Latent Hierarchical Causal Structure Discovery with Rank Constraintscited by 50
  8. 2020CausalVAE: Structured Causal Disentanglement in Variational Autoencodercited by 47
  9. 2020CRAFT: A Benchmark for Causal Reasoning About Forces and inTeractionscited by 44
  10. 2023Subspace Identification for Multi-Source Domain Adaptationcited by 36
  11. 2024A Sparsity Principle for Partially Observable Causal Representation Learningcited by 34
  12. 2013A Discovery Algorithm for Directed Cyclic Graphscited by 33
  13. 2024Cross-Entropy Is All You Need To Invert the Data Generating Processcited by 30
  14. 2022Partial Disentanglement via Mechanism Sparsitycited by 29
  15. 2013Theory Refinement on Bayesian Networkscited by 28
  16. 2025CausalRivers -- Scaling up benchmarking of causal discovery for real-world time-seriescited by 24
  17. 2024CaRiNG: Learning Temporal Causal Representation under Non-Invertible Generation Processcited by 20
  18. 2024On the Identification of Temporally Causal Representation with Instantaneous Dependencecited by 15
  19. 2024Smoke and Mirrors in Causal Downstream Taskscited by 12
  20. 2025CAUSAL3D: A Comprehensive Benchmark for Causal Learning from Visual Datacited by 7
  21. 2025Sanity Checking Causal Representation Learning on a Simple Real-World Systemcited by 5

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2026: 3.

3 papers.

  1. 2026Who Guards the Guardians? The Challenges of Evaluating Identifiability of Learned Representationscited by 0
  2. 2026A Dialogue between Causal and Traditional Representation Learning: Toward Mutual Benefits in a Unified Formulationcited by 0
  3. 2026MoVA: Learning Asymmetric Dual Projections for Modular Long Video-Text Alignmentcited by 0