- Rating
- 1
- Confidence
- 4
- Informativeness
- 1
- Presentation
- 2
- Soundness
- 1
Read review
Paper summary
The paper estimates the Tian-Pearl bounds on the conditional probability of necessity and sufficiency, PNS(z) = P(Y₁ = 1, Y₀ = 0 | Z = z), from observational and randomized data with binary treatment and outcome. A single neural network with three output heads estimates the observational probabilities P(X = x, Y = y | Z = z) and the experimental probabilities P(Y = 1 | do(X = x), Z = z). The experimental outputs are parameterized so that they always satisfy the compatibility constraints between the two data sources (Eq. 5), for example P(X = 1, Y = 1 | Z = z) ≤ P(Y = 1 | do(X = 1), Z = z) ≤ 1 − P(X = 1, Y = 0 | Z = z). The bound terms are then combined with the precision correction of Chernozhukov, Lee and Rosen (2013), with standard errors and critical values taken from Epistemic Neural Network draws.
Critical strengths
Easy to understand.
Critical weaknesses
W1. The ENN spread is not a standard error. The precision correction requires standard errors and a critical value from the sampling distribution of the term estimators. Here they come from the spread of ENN draws, which is an approximate posterior spread whose size is set by the prior scale, a tuning parameter selected by validation loss and calibration error (1.25 in the benchmarks and 2 in the case study, Appendix E). These standard errors are therefore not accurate, and the resulting correction is not theoretically sound.
W2. The coverage guarantee is for a different method. The proposed method takes its standard errors from ENN draws, but the coverage theorem (Theorem B.1, Corollary B.4) is proved for an influence-function multiplier bootstrap, which the experiments use only as a baseline (MB Full), as Section 3.2 itself states. That theorem also assumes the network is a correctly specified parametric model with an invertible Hessian, which overparameterized networks do not satisfy, so no method in the experiments has a coverage guarantee.
Missing comparisons and literature. The experiments compare only S- and T-learner plug-ins and the authors' own variants. Existing methods for the same quantities are cited but not compared: debiased and smoothed inference for bounds with guarantees (Semenova, 2023; Whitehouse et al., 2026), machine-learning estimation of probabilities of causation (Wang and Li, 2025), and bounds on the fraction harmed (Kallus, 2022). Closely related work is not cited: the nonparametric efficient estimator for the probability of causation (Cuellar and Kennedy, JRSS-A 2020), debiased estimation of covariate-assisted bounds (Levis et al., 2023), and partial counterfactual identification from observational and experimental data, including PNS (Mueller, Li and Pearl, IJCAI 2022; Zhang, Tian and Bareinboim, ICML 2022; Xia, Pan and Bareinboim, ICLR 2023).
Questions for authors
- Do the ENN standard errors match the standard deviation of the estimates across the 1,000 replicates?
- How do coverage and width change with the prior scale of the ENN?
Validation Protocol, Crossed Intervals, and Case-Study Stability (W3, Q4)
Anchored,MB Full,MB Last Lyr., andAnchored EpiNet(0% crossed intervals).S-learner(2.36% of test points) andT-learner(2.56%). In evaluation (Appendix E.6), coverage checks 1[ℓi≤yi≤ui] and 1[ℓi≤L(zi) and U(zi)≤ui] directly, so any crossed interval (ℓi>ui) automatically evaluates to 0 (non-coverage). For the Winkler interval score ISi=(ui−ℓi)+α2(ℓi−yi)++α2(yi−ui)+, endpoints are passed unchanged without reordering, so any crossed interval incurs a net penalty of at least (α2−1)(ℓi−ui)=39(ℓi−ui) at α=0.05. We will state this explicitly in Appendix E.6 and report the crossed-interval rates alongside% Valid.The refit means differ from Figure 4 by at most 0.4 percentage points with standard deviations of only 0.8–1.3 percentage points, demonstrating that aggregate within-bin PNS non-overlap rates are stable across data splits and training seeds.