mlsanityData to 7 Oct 2026

Beyond Coefficients: Forecast-Necessity Testing for Interpretable Causal Discovery in Nonlinear Time-Series Models

arXiv 2604.18751 ↗First posted 20 Apr 2026cs.LG cs.AI stat.MEcited by 0

Not on the map. Why?

Abstract (arXiv)

Nonlinear machine-learning models are increasingly used to discover causal relationships in time-series data, yet the interpretation of their outputs remains poorly understood. In particular, causal scores produced by regularized neural autoregressive models are often treated as analogues of regression coefficients, leading to misleading claims of statistical significance. In this paper, we argue that causal relevance in nonlinear time-series models should be evaluated through forecast necessity rather than coefficient magnitude, and we present a practical evaluation procedure for doing so. We present an interpretable evaluation framework based on systematic edge ablation and forecast comparison, which tests whether a candidate causal relationship is required for accurate prediction. Using Neural Additive Vector Autoregression as a case study model, we apply this framework to a real-world case study of democratic development, modeled as a multivariate time series of panel data - democracy indicators across 139 countries. We show that relationships with similar causal scores can differ dramatically in their predictive necessity due to redundancy, temporal persistence, and regime-specific effects. Our results demonstrate how forecast-necessity testing supports more reliable causal reasoning in applied AI systems and provides practical guidance for interpreting nonlinear time-series models in high-stakes domains.

Key facts. No citing papers in the map yet (data through 7 Oct 2026). Builds on 30 of its 65 references that are in the map. First posted on arXiv 20 Apr 2026.

Builds on

30 of 65 references.

References · 30

  1. 2017Proximal Policy Optimization Algorithmscited by 14,441
  2. 2017Towards Deep Learning Models Resistant to Adversarial Attackscited by 6,326
  3. 2015High-Dimensional Continuous Control Using Generalized Advantage Estimationcited by 1,916
  4. 2018Spectral Normalization for Generative Adversarial Networkscited by 1,618
  5. 2017Domain Randomization for Transferring Deep Neural Networks from Simulation to the Real Worldcited by 1,401
  6. 2020Language (Technology) is Power: A Critical Survey of "Bias" in NLPcited by 951
  7. 2021Isaac Gym: High Performance GPU-Based Physics Simulation For Robot Learningcited by 882
  8. 2017Sim-to-Real Transfer of Robotic Control with Dynamics Randomizationcited by 616
  9. 2018Lipschitz regularity of deep neural networks: analysis and efficient estimationcited by 413
  10. 2020Robust Deep Reinforcement Learning against Adversarial Perturbations on State Observationscited by 235
  11. 2019Adversarial Policies: Attacking Deep Reinforcement Learningcited by 209
  12. 2019Action Robust Reinforcement Learning and Applications in Continuous Controlcited by 167
18 more
  1. 2021Robust Reinforcement Learning on State Observations with Learned Optimal Adversarycited by 133
  2. 2022Walk These Ways: Tuning Robot Control for Generalization with Multiplicity of Behaviorcited by 113
  3. 2020Exactly Computing the Local Lipschitz Constant of ReLU Networkscited by 102
  4. 2021Online Robust Reinforcement Learning with Model Uncertaintycited by 95
  5. 2022Learning Smooth Neural Functions via Lipschitz Regularizationcited by 86
  6. 2018Reinforcement Learning with Perturbed Rewardscited by 76
  7. 2022Policy Gradient Method For Robust Reinforcement Learningcited by 67
  8. 2020Challenges and Countermeasures for Adversarial Attacks on Deep Reinforcement Learningcited by 61
  9. 2019Wasserstein Robust Reinforcement Learningcited by 50
  10. 2022Efficiently Computing Local Lipschitz Constants of Neural Networks via Bound Propagationcited by 45
  11. 2020Robustifying Reinforcement Learning Agents via Action Space Adversarial Trainingcited by 30
  12. 2022L2C2: Locally Lipschitz Continuous Constraint towards Stable and Smooth Reinforcement Learningcited by 22
  13. 2021Learning Robust Policy against Disturbance in Transition Dynamics via State-Conservative Policy Optimizationcited by 18
  14. 2020Lipschitz Bounds and Provably Robust Training by Laplacian Smoothingcited by 17
  15. 2024Robust Deep Reinforcement Learning Through Adversarial Attacks and Training : A Surveycited by 13
  16. 2022Robust Reinforcement Learning in Continuous Control Tasks with Uncertainty Set Regularizationcited by 7
  17. 2019A Transition-Aware Method for the Simulation of Compliant Contact with Regularized Frictioncited by 3
  18. 2022Interval Observer Synthesis for Locally Lipschitz Nonlinear Dynamical Systems via Mixed-Monotone Decompositionscited by 2

Built on by

Not cited yet.

Similar, not cited

Close in wording, no citation either way

  1. 2026From Causal Discovery to Dynamic Causal Inference in Neural Time Series
  2. 2025Augur: Modeling Covariate Causal Associations in Time Series via Large Language Models
  3. 2024Explaining Time Series Classification Predictions via Causal Attributions
  4. 2025TimeMKG: Knowledge-Infused Causal Reasoning for Multivariate Time Series Modeling
  5. 2025Causality can systematically address the monsters under the bench(marks)
  6. 2023Causal Reasoning in the Presence of Latent Confounders via Neural ADMG Learning
  7. 2025The Third Pillar of Causal Analysis? A Measurement Perspective on Causal Representations
  8. 2024A Dynamical View of the Question of Why