# DoTime: A Synthetic Benchmark Generator for Interventional and Counterfactual Time Series

arXiv: https://arxiv.org/abs/2607.27263 · HTML: https://mlsanity.com/paper/2607.27263/

## Key facts

- Cited by 2 papers in the map (data through 14 Oct 2026).
- Builds on 10 of its 27 references that are in the map.
- First posted on arXiv 29 Jul 2026.

## Builds on (most cited first, 10 of 10)

- [Datasheets for Datasets](https://mlsanity.com/paper/1803.09010/) (2018)
- [Chronos-2: From Univariate to Universal Forecasting](https://mlsanity.com/paper/2510.15821/) (2025)
- [Croissant: A Metadata Format for ML-Ready Datasets](https://mlsanity.com/paper/2403.19546/) (2024)
- [Do-PFN: In-Context Learning for Causal Effect Estimation](https://mlsanity.com/paper/2506.06039/) (2025)
- [CausalTime: Realistically Generated Time-series for Benchmarking of Causal Discovery](https://mlsanity.com/paper/2310.01753/) (2023)
- [Atlantic Causal Inference Conference (ACIC) Data Analysis Challenge 2017](https://mlsanity.com/paper/1905.09515/) (2019)
- [Interventional Time Series Priors for Causal Foundation Models](https://mlsanity.com/paper/2603.11090/) (2026)
- [Position: Causal Machine Learning Requires Rigorous Synthetic Experiments for Broader Adoption](https://mlsanity.com/paper/2508.08883/) (2025)
- [Towards Continuous-time Causal Foundation Models](https://mlsanity.com/paper/2605.28880/) (2026)
- [Causal Graphs, Markov Properties and Do-calculus for Stochastic Differential Equations](https://mlsanity.com/paper/2607.12140/) (2026)

## Built on by (most cited first, 2 of 2)

- [What You Observe Determines How You Identify Causal Effects: Evaluating Causal Models across Observational Views](https://mlsanity.com/paper/2609.36881/) (2026)
- [On the Divergence of Accuracy and Mechanism Consistency in Time Series World Models](https://mlsanity.com/paper/2610.01842/) (2026)

## Abstract (arXiv)

Most benchmarks for causal inference over time series are observational, small, or domain-specific, leaving interventional and counterfactual estimation under-served exactly where it matters most, such as in healthcare, policy evaluation, and climate science. We introduce DoTime, an open, scalable, and theoretically grounded generator of multivariate temporal structural causal models (TSCMs) with interventions, released as the codedotime PyPI package together with four frozen evaluation suites. Beyond existing work, it adds capabilities absent from prior generators: continuous-time intervention windows, counterfactual sampling modes with a positivity guard, regime-switching SCMs as a strict generalization of interrupted time series, non-stationary dynamics by construction with switching SCM parameters, and deterministic ramp and sinusoidal intervention profiles that place trends and structural breaks inside the evaluation window. Moreover, it demonstrates the suitability of the generator as a prior for a causal foundation model reference implementation. The released suites span a training-scale snapshot of 100,000 trajectories and eight named identification structures, each with exact ground truth: paired interventional trajectories from the same SCM throughout, and shared-noise counterfactuals in the continuous-time suite. We ship reference baseline implementations with an evaluation harness, and pose a falsifiable claim: interventional training buys a measurable direction-accuracy advantage over an observational model of identical capacity. It is tested across three training seeds per arm. Under structure-matched evaluation on held-out episodes, the interventional prior-fitted network's (PFN) gap is positive in every structure, trajectory length, and seed tested.