mlsanityData to 14 Oct 2026

CausalArena: Benchmarking Causal Discovery in the Foundation Model Era

arXiv 2609.11897 ↗First posted 10 Sep 2026cs.LGcited by 0

On the map: LLM reasoning → Causal reasoning in LLMs

Abstract (arXiv)

Causal discovery aims to uncover causal structures from data and is fundamental to scientific reasoning and intervention-based decision making. Its evaluation relies heavily on structural causal models (SCMs), which specify a causal graph together with the mechanisms that generate data, yet existing studies differ substantially in graph families, mechanisms, and evaluation protocols. The emergence of causal discovery foundation models (CDFMs) further complicates evaluation: performance may reflect not only causal discovery ability, but also overlap between pretraining environments and test SCMs, making results on fixed synthetic benchmarks difficult to interpret. We introduce CausalArena, a unified and evolvable benchmark for causal discovery under a common protocol. Synthetic SCMs supply controlled breadth over structures and mechanisms; semantic operational SCMs provide human-auditable, semantically grounded environments beyond standard synthetic generators; and formula-grounded SCMs test discovery under explicit scientific mechanisms. Public real-world datasets provide an additional external-validity check. Experiments across classical, neural, and pretrained methods reveal substantial ranking shifts across SCM families and protocols, showing that strong performance in one benchmark regime does not reliably transfer to others. These results highlight benchmark diversity and pretraining--evaluation overlap as central challenges for evaluating causal discovery in the foundation model era.

Key facts. No citing papers in the map yet (data through 14 Oct 2026). Placed in Causal reasoning in LLMs, one of 17 directions in LLM reasoning. Builds on 29 of its 46 references that are in the map. First posted on arXiv 10 Sep 2026.

Builds on

29 of 46 references.

References · 29

  1. 2018DAGs with NO TEARS: Continuous Optimization for Structure Learningcited by 752
  2. 2020Differentiable Causal Discovery from Interventional Datacited by 218
  3. 2019Learning Sparse Nonparametric DAGscited by 207
  4. 2013CAM: Causal additive models, high-dimensional order search and penalized regressioncited by 178
  5. 2021Beware of the Simulated DAG! Causal Discovery Benchmarks May Be Easy To Gamecited by 161
  6. 2022DAGMA: Learning DAGs via M-matrices and a Log-Determinant Acyclicity Characterizationcited by 140
  7. 2022Deep End-to-end Causal Inferencecited by 95
  8. 2022Amortized Inference for Causal Structure Learningcited by 90
  9. 2017Permutation-based Causal Inference Algorithms with Interventionscited by 78
  10. 2023Scalable Causal Discovery with Score Matchingcited by 45
  11. 2023CausalTime: Realistically Generated Time-series for Benchmarking of Causal Discoverycited by 36
  12. 2023Stable Differentiable Causal Discoverycited by 31
17 more
  1. 2025CausalRivers -- Scaling up benchmarking of causal discovery for real-world time-seriescited by 26
  2. 2024Large Language Models for Causal Discovery: Current Landscape and Future Directionscited by 25
  3. 2024Sample, estimate, aggregate: A recipe for causal discovery foundation modelscited by 21
  4. 2023NODAGS-Flow: Nonlinear Cyclic Causal Structure Learningcited by 20
  5. 2023textttcausalAssembly: Generating Realistic Production Data for Benchmarking Causal Discoverycited by 18
  6. 2023The PetShop Dataset -- Finding Causes of Performance Issues across Microservicescited by 14
  7. 2025When Selection Meets Intervention: Additional Complexities in Causal Discoverycited by 10
  8. 2025TimeGraph: Synthetic Benchmark Datasets for Robust Time-Series Causal Discoverycited by 10
  9. 2025Unitless Unrestricted Markov-Consistent SCM Generation: Better Benchmark Datasets for Causal Discoverycited by 8
  10. 2025CausalDynamics: A large-scale benchmark for structural discovery of dynamical causal modelscited by 8
  11. 2026CauScale: Neural Causal Discovery at Scalecited by 6
  12. 2025Bayesian Network Structure Discovery Using Large Language Modelscited by 5
  13. 2026Arrow: A Foundation Model for Causal Discoverycited by 5
  14. 2026TabCausal: Pretraining Across Causal Environments for Tabular Causal Discoverycited by 4
  15. 2026FoundCause: Causal Discovery with Latent Confounders from Observational Datacited by 3
  16. 2025CausalProfiler: Generating Synthetic Benchmarks for Rigorous and Transparent Evaluation of Causal Machine Learningcited by 2
  17. 2026CDFM: Towards a General-Purpose Causal Discovery Foundation Modelcited by 2

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2025LLM Cannot Discover Causality, and Should Be Restricted to Non-Decisional Support in Causal Discovery
  2. 2026Leveraging Large Language Models for Causal Discovery: a Constraint-based, Argumentation-driven Approach
  3. 2025Causal-Copilot: An Autonomous Causal Analysis Agent
  4. 2025Causal Reasoning in Pieces: Modular In-Context Learning for Causal Discovery
  5. 2024LLM-initialized Differentiable Causal Discovery
  6. 2026CausalDS: Benchmarking Causal Reasoning in Data-Science Agents
  7. 2026CausalReasoningBenchmark: A Real-World Benchmark for Disentangled Evaluation of Causal Identification and Estimation
  8. 2026Searching for What Matters: Open-Ended Discovery of Multiple Hypotheses with Agentic Data Exploration