mlsanityData to 7 Oct 2026

Interventional Time Series Priors for Causal Foundation Models

arXiv 2603.11090 ↗First posted 11 Mar 2026cs.LG stat.MEcited by 6

Not on the map. Why?

Abstract (arXiv)

Prior-data fitted networks (PFNs) have emerged as powerful foundation models for tabular causal inference, yet their extension to time series remains limited by the absence of synthetic data generators that provide interventional targets. Existing time series benchmarks generate observational data with ground-truth causal graphs but lack the interventional data required for training causal foundation models. To address this, we propose CausalTimePrior, a principled framework for generating synthetic temporal structural causal models (TSCMs) with paired observational and interventional time series. Our prior supports configurable causal graph structures, nonlinear autoregressive mechanisms, regime-switching dynamics, and multiple intervention types (hard, soft, time-varying). We demonstrate that PFNs trained on CausalTimePrior can perform in-context causal effect estimation on held-out TSCMs, establishing a pathway toward foundation models for time series causal inference.

Key facts. Cited by 6 papers in the map (data through 7 Oct 2026). Builds on 23 of its 34 references that are in the map. First posted on arXiv 11 Mar 2026.

Builds on

23 of 34 references.

References · 23

  1. 2018Neural Ordinary Differential Equationscited by 3,733
  2. 2022TabPFN: A Transformer That Solves Small Tabular Classification Problems in a Secondcited by 643
  3. 2021Transformers Can Do Bayesian Inferencecited by 338
  4. 2013Causal Discovery with Continuous Additive Noise Modelscited by 252
  5. 2021Neural SDEs as Infinite-Dimensional GANscited by 159
  6. 2020Estimating Counterfactual Treatment Outcomes over Time Through Adversarially Balanced Representationscited by 154
  7. 2020Discovering contemporaneous and lagged causal relations in autocorrelated nonlinear time series datasetscited by 136
  8. 2022Causal Transformer for Estimating Counterfactual Outcomescited by 121
  9. 2023ForecastPFN: Synthetically-Trained Zero-Shot Forecastingcited by 102
  10. 2025Do-PFN: In-Context Learning for Causal Effect Estimationcited by 46
  11. 2023CausalTime: Realistically Generated Time-series for Benchmarking of Causal Discoverycited by 35
  12. 2025Foundation Models for Causal Inference via Prior-Data Fitted Networkscited by 30
11 more
  1. 2025TimePFN: Effective Multivariate Time Series Forecasting with Synthetic Datacited by 27
  2. 2025CausalRivers -- Scaling up benchmarking of causal discovery for real-world time-seriescited by 24
  3. 2024Signature Kernel Conditional Independence Tests in Causal Discovery for Stochastic Processescited by 18
  4. 2025TempoPFN: Synthetic Pre-training of Linear RNNs for Zero-shot Time Series Forecastingcited by 16
  5. 2024RealTCD: Temporal Causal Discovery from Interventional Data with Large Language Modelcited by 13
  6. 2024Dynamic Structural Causal Modelscited by 13
  7. 2023On the Identifiability of Switching Dynamical Systemscited by 9
  8. 2025TimeGraph: Synthetic Benchmark Datasets for Robust Time-Series Causal Discoverycited by 9
  9. 2025Causal Time Series Generation via Diffusion Modelscited by 7
  10. 2024CAnDOIT: Causal Discovery with Observational and Interventional Data from Time-Seriescited by 3
  11. 2025Towards Causal Market Simulatorscited by 2

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2026: 6.

6 papers.

  1. 2026Towards Continuous-time Causal Foundation Modelscited by 3
  2. 2026DoTime: A Synthetic Benchmark Generator for Interventional and Counterfactual Time Seriescited by 2
  3. 2026Causal Longitudinal Prior-Fitted Networks for Counterfactual Outcome Predictioncited by 1
  4. 2026Temporal Causal Prior-Data Fitted Networks for Panel Data with Learned Reliability Signalscited by 0
  5. 2026AutoCause: A Python framework that automates expert decisions in environmental time-series causal discoverycited by 0
  6. 2026Causal Foundation Modelscited by 0