- Rating
- 2(+1)
- Confidence
- 3
- Informativeness
- 1
- Presentation
- 2(+1)
- Soundness
- 2(+1)
Read review
Paper summary
The paper studies linear latent causal models with observed instruments W and measurements X with unknown instrument targets, and sparse loadings that can include cross-loadings. A reverse-topological peeling procedure separates these directions into a latent DAG and loading matrix. The paper gives necessary-and-sufficient conditions for this population algorithm and a finite-sample clustering and trimmed-regression implementation. Empirical experiments include a simulation dataset with additional scaling and assumption stresses, dimension-selection experiments, and a three-gene masked OneK1K downstream genetic application.
Critical strengths
Reporting weak-IV failures and incomplete recovery provides a useful characterization of the method’s limitations.
Critical weaknesses
- The IV assumptions are difficult to justify in real-world data. Identification requires single-target instruments, coverage of every latent variable, and exclusion of pathways bypassing the modeled latent system. How can these assumptions be justified?
- The comparative evaluation is limited to basic representation-plus-GES pipelines. No existing interventional CRL method is evaluated.
- The empirical results are not comprehensive enough. The baselines (PCA, sparse-varimax and IV-SVD + GES) estimate a rotated basis and therefore fail by construction. The comparison is not convincing to show the advantage of the method.
Questions for authors
- The method requires a complete collection of valid, single-target instruments for unobserved latent variables, but provides limited guidance on how such instruments can be obtained or justified in practice.
- In real world datasets OneK1K genetic application, how to justify the model assumptions hold ?
- Do prop. 1 and theorems 2-3 allow an arbitrary , i.e., hidden confounding among the latent variables?
- How were s=4, q=2 and tau=0.06 chosen for OneK1K, and how stable are the conclusions to these choices? Sensitivity analysis is needed 5.The current comparisons largely rely on factor extraction followed by GES. The authors should compare against strong published latent-variable discovery methods, such as RLCD, LGES, FOFC/MIMBuild in a fair setting.
5 comments
Q4: choice and sensitivity of s, q, and tau
With three latent nodes, the largest possible outdegree is . Hence imposes no additional degree restriction. The tolerance came from simulations as it works reasonably well across various settings. The choice of was a difficult one: too small, the algorithm will exit (step 4) without completion; too big, it can absorb latent causal effects into the loading matrix and return an empty graph. One could in principle tune this parameter to achieve the desired sparsity of the latent graph. That said, requiring the user to specify is indeed a practical limitation of the original peeling algorithm. We now provide two alternatives that avoid these inputs, Orthogonal LP/refit and Orthogonal robust-order, that do not take as inputs.
- Orthogonal LP/refit: estimate IV total-effect directions in whitened coordinates; propose sparse directions with linear programs; select one acyclic order; and refit/prune edges by robust regression that propagates first-stage uncertainty.
- Orthogonal robust-order: use the same first stage and robust regression, but search over every candidate child set and order by dynamic programming. The implementation is intended for small latent dimension ().
For these two variants, the original P1-P3 are replaced by the identification condition on the loading matrix
Condition (S). For every vector with ,
where is the th loading column. Thus mixing two or more loading columns must produce more nonzeros than every participating column. Under full-rank loadings and the IV assumptions, (S) makes the minimum-total-support basis unique up to scaling and permutation; acyclicity then identifies from up to latent scaling and relabeling. We retain this result alongside the original P1–P3 characterization of population peeling.
We reran the exact original main simulation grid and seeds, without changing its generating models: , , IV strength , measurement SNR , and one/two IVs per latent, with 100 replicates in each of 24 cells. The original peeler receives its original ; the new variants receive neither value. All original datasets are retained: none was filtered to satisfy (S), and we have not certified (S) for every generated loading matrix.
| Method | TPR | FDR | Complete fits |
|---|---|---|---|
| Original peeler, | 50.46% | 13.99% | 1,123/2,400 |
| Original peeler, | 57.43% | 14.46% | 1,440/2,400 |
| Original peeler, | 57.96% | 14.95% | 1,685/2,400 |
| Orthogonal LP/refit | 84.57% | 8.41% | 2,400/2,400 |
| Orthogonal robust-order | 86.67% | 10.25% | 2,400/2,400 |
TPR is total true positives divided by all 5,274 true directed edges across the scheduled datasets. FDR is mean dataset-level , with empty predictions contributing zero. Partial peeler outputs are retained and scored against the full true graph. Latent labels are aligned using IV-assignment overlap, without choosing labels to maximize graph agreement. Monte Carlo SEs for TPR/FDR, in percentage points, are for LP/refit, for robust-order, and for the original peeler.
In addition to being free of , both variants increase TPR and reduce FDR. For example, At , LP/refit reaches TPR/FDR and robust-order , compared with for the original peeler.
W2: comparison with interventional CRL
We agree and have evaluated Squires et al. [3] and LSCALE-I [9] using their public numerical implementations on the same 2,400 original IV datasets, alongside the original peeling algorithm and both new IV-CRL variants. All methods receive the common supplied dimension ; IVs define the environments for the interventional methods.
We first regress jointly on all instruments using the original observations. We normalize and cluster the estimated IV response columns into groups, treating opposite directions as equivalent and using no true target labels. Within each learned group, we select the IV with the largest estimated response-column norm.
Let . The baseline consists of the first observations in their original i.i.d. order. For each selected IV, we take the observations, from all rows, whose IV values are closest to its sample median. Their observed values form a central-half IV environment. For the independent Gaussian IVs used here, this restriction changes the selected IV's population variance relative to the baseline distribution. We center each environment separately and supply the resulting environments to both methods. Each has observations; environments overlap and contain at most distinct subjects in total, so reused rows are not counted as independent additional observations.
Squires uses its public partial-order routine and estimated covariance matrices; LSCALE-I uses estimated Gaussian scores. The primary numerical thresholds and all sensitivity values are recorded before final aggregation.
We report direct edges and ancestral relations separately, since the targets differ:
| Method/output at its primary threshold | Evaluation target | TPR | FDR |
|---|---|---|---|
| Original peeler, (our method) | Direct edges | 57.43% | 14.46% |
| Orthogonal LP/refit (our method) | Direct edges | 84.57% | 8.41% |
| Orthogonal robust-order (our method) | Direct edges | 86.67% | 10.25% |
| Squires, coefficient support, cutoff | Direct edges | 7.38% | 60.97% |
| Squires, native partial-order output | Ancestral relations | 46.88% | 86.96% |
| LSCALE-I, soft output, threshold | Ancestral relations | 0.00% | 0.00% |
All these runs returned outputs; the peeler completed all nodes in 1,440/2,400 datasets, and its partial outputs remain included. Native ancestral outputs are compared with the true transitive closure, not with the direct DAG. LSCALE-I's empty outputs have zero FDR and zero recovery. This result is threshold-dependent: its soft branch at gives ancestral TPR with FDR, and at gives . Squires' direct cutoffs and give and , respectively. The separately labeled LSCALE-I hard refinement, applied to these soft IV environments, is misspecified and is not treated as a valid hard-intervention comparison.
Scope of this comparison. An additive IV changes a targeted disturbance distribution while preserving the latent parent coefficients. Such noise-only changes fall outside Squires' relevant genericity condition and, for nonzero parent effects, LSCALE-I's score-change condition [3,9]. In particular, a conditional log-density ratio has proportional target and parent derivatives. Noisy measurement cells also depart from deterministic mixing, and Gaussian score estimation is approximate after IV stratification. Therefore, the proposed methods outperforming them is not surprising.
Thank you for your valuable comments.
W1 and Q1: obtaining valid single-target instruments with full coverage
We agree that IV validity is difficult to establish from observed data alone and that obtaining valid single-target instruments covering every latent variable requires substantive justification. This is however not unique to this work but rather reflects a very broader inherent challenge in causal representation learning: many key identifying assumptions concern unobserved variables or mechanisms. Examples include independence of latent structural disturbances in non-Gaussian approaches [1], restrictions on which latent groups directly generate particular measurements in GIN [2], and interventions that change a single latent mechanism while preserving the measurement map in interventional CRL [3]. Even though some of these assumptions have testable implications that allow certain violations to be detected, satisfying those implications, however, does not establish that the underlying assumptions hold.
Despite these challenges, IV methods are widely used in econometrics, including studies of supply and demand and returns to schooling [4], and in genetics through Mendelian randomization [5]. Their use rests on scientific or institutional arguments for instrument validity, supported by empirical diagnostics and sensitivity analyses where available. The proposed IV-CRL brings this established source of causal information to representation learning, allowing Gaussian and correlated structural disturbances under its exogeneity and structural assumptions.
Our assumptions on instruments for latent variables are analogous to the relevance, exogeneity, and exclusion assumptions used in IV-based causal discovery with observed variables. This parallels non-Gaussian CRL methods [1], which extend the independent, non-Gaussian disturbance assumptions of causal discovery with observed variables [6] to latent systems. In IV-CRL, single targeting expresses exclusion relative to an instrument's designated latent target: its effects on other latents and measurements must pass through that target. Full coverage supplies relevance across the latent system and, under single targeting, ensures rank(Γ)=d. A comparable coverage requirement appears in interventional CRL, where full identification may require an intervention on every latent variable [3].
We envisage applications to a focused set of latent traits or regulators for which external knowledge supports plausible instruments. For example, in genetically informed psychological measurement, W contains candidate SNPs, Z contains latent traits such as anxiety and depression, and X contains questionnaire measurements [7,8]. A valid single-target SNP directly affects one trait; its effects may propagate to other traits and measurements through that target.
In genetics, external association evidence, cis localization, conditional relevance screening, LD filtering, and ancestry adjustment provide practical starting points. Collecting multiple plausible candidates per target can improve the prospect of coverage; our model also already permits irrelevant candidates.
Q2: justification in OneK1K
As discussed in our response to W1/Q1, IV validity is generally difficult to establish from observed data alone. This reflects the broader challenge of assessing identifying assumptions about unobserved variables and mechanisms in CRL. The same limitation applies to OneK1K: the cis-associated SNPs provide scientifically motivated candidate instruments, while exogeneity and exclusion remain substantive assumptions. Our interpretation of the application is conditional on these assumptions.
Q3: arbitrary covariance among latent structural disturbances
Yes. For the IV response matrix Θ=Cov(X,W)Cov(W)−1, the moment assumptions give
Θ=MΓ
under Cov(ε,W)=Cov(η,W)=0, regardless of within-ε or within-η covariance. Both the original population-peeling guarantee under P1–P3 and the complementary sparsest-basis identification result under (S), defined in Q4 below, operate on this identified M and do not need diagonal disturbance covariance.