Pith. sign in

REVIEW 4 major objections 4 minor 70 references

Differentiable Causal Discovery For Latent Hierarchical Causal Models

T0 review · 4 major / 4 minor · reviewed 2026-08-12 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read The paper proves that nonlinear latent hierarchical causal models are identifiable from observational data without assuming latent variables or noise are deterministic functions of measurements, and obtains the graph via a Jacobian rank…

desk verdict The paper's central theorem is false as stated, and the proof is circular; the algorithmic contribution is real but the theory cannot be used. read the letter →

arxiv 2411.19556 v1 pith:XARLJFNO submitted 2024-11-29 cs.LG

Parjanya Prashant , Ignavier Ng , Kun Zhang , Biwei Huang This is my paper · ORCID
classification cs.LG
keywords latenthierarchicalcausalmodelsdiscoveryidentifiabilityJacobianranknonlinearSEMvariationalautoencoderdifferentiableoptimizationd-separation
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

This paper tries to show that the causal structure of nonlinear hierarchical models with unobserved variables can be recovered from purely observational data, and that the earlier deterministic-function assumption (latents and noise are invertible functions of the measurements) can be dropped. The load-bearing criterion is a new rank test: the rank of the Jacobian of the conditional expectation E[y|x] equals the number of latent variables in the smallest set that d-separates x from y. From this test the paper proves the whole graph is identifiable up to relabeling of the latents, and it builds a differentiable algorithm—a VAE with Gumbel-softmax adjacency masks and a two-pure-children constraint—that learns the graph in one continuous optimization run. The method is demonstrated on synthetic graphs and on image data, where the learned hierarchy is interpretable and transfers across distribution shifts.

What carries the argument

The load-bearing object is the Jacobian rank equality rank(∂E[y|x]/∂x) = size of the smallest d-separating latent set (Theorem 1). It converts a graphical notion, d-separation, into a continuous, data-observable quantity, which is what allows discrete structure search to be replaced by differentiable optimization. Supporting machinery: Theorem 2 lets measured pure descendants stand in for latent sets; Lemmas 1–3 turn rank equalities into rules for identifying pure-child groups, the parent set of non-pure children, and parentless measured variables; and the algorithm uses a block upper-triangular Gumbel-softmax adjacency matrix inside a VAE whose decoder respects the SEM, with a constraint that each latent row has at least two pure children.

What would settle it

Simulate a nonlinear hierarchical model satisfying the paper's Condition 1 with a measured variable that has two latent parents (as in the paper's Figure 1), and compute the rank of the Jacobian of E[Y|X] for X equal to that single variable and Y any other measured set. The minimal d-separating set has size 2, while the scalar-input Jacobian has rank at most 1; if the paper's lemmas rely on these two numbers being equal, they are falsified on this graph.

Watch

Extended reading notes

Core claim

The central claim is Theorem 1: under a generalized faithfulness condition and a differentiability condition, for any two measured sets X and Y with r < |X|, |Y|, the rank of J_f = ∂E[y|x]/∂x equals the size of the smallest set of latent variables that d-separates X from Y. This generalizes the rank-deficiency constraints used in linear latent-variable models to the nonlinear case and removes the deterministic-function requirement of Kong et al. (2023). The paper further shows that pure measured descendants can stand in for latent sets (Theorem 2), which allows a recursive procedure: identify the lowest latent layer from the measured variables using three rank-based lemmas, then lift to the next layer using the identified pure children as surrogates. The result (Theorem 3) is that the binary adjacency matrix M of the hierarchical graph is identifiable up to permutation of the latent variables.

Load-bearing premise

The proof assumes the smallest set of hidden variables that separates two measured groups is never larger than the smaller of the two groups, yet the paper's own conditions allow a single measured variable to have two latent parents, where that assumption fails for singleton groups.

Editorial extensions

If this is right

  • Nonlinear latent hierarchical causal models become identifiable from observational data without requiring latent variables or noise to be deterministic functions of the measurements.
  • The Jacobian rank of E[y|x] provides a general, distribution-level test for how many hidden variables separate any two measured sets, extending rank constraints beyond linear models.
  • The differentiable VAE-based algorithm learns the graph in one continuous optimization run, avoiding the iterative training of many generative models and scaling to hundreds of latent variables.
  • On image data, the recovered hierarchy is interpretable and its representations transfer across distribution shifts (CMNIST, CelebA), supporting the use of causal latent structure for domain adaptation.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • The Jacobian-rank criterion could likely be repurposed as a nonparametric test for latent confounding between arbitrary variable sets, whenever each side has enough measured variables to carry the dimension.
  • The two-pure-children condition is probably not the exact identifiability boundary; nonlinear mechanisms with sparse footprints may still permit identification under additional smoothness or noise assumptions.
  • The need to fix the number of latent layers in advance suggests a practical extension: run the differentiable search with varying layer counts and select by held-out likelihood, effectively performing structure selection over depth.
  • Because the rank test is based on conditional expectations, finite-sample estimation error in the regression function will bias rank estimates downward; a bias-correction or a hypothesis-testing version of the rank equality would be a natural next step.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

4 major / 4 minor

Summary. The paper proposes a theoretical framework for identifying nonlinear latent hierarchical causal models from observational data, claiming to relax the deterministic-function assumption of Kong et al. (2023). The central result, Theorem 1, states that the rank of the Jacobian of the conditional expectation E[y|x] equals the size of the smallest latent set that d-separates the measured sets X and Y. Building on this, the paper derives an identifiability theorem (Theorem 3) and presents a differentiable causal discovery algorithm based on a VAE with Gumbel-softmax masks and structural constraints. Experiments on synthetic graphs and image datasets are reported to support the method's accuracy and scalability.

Significance. If the main theorem were correct, the paper would be a substantial step forward: it would provide a continuous, scalable rank-based criterion for latent hierarchical structure without requiring linearity or deterministic invertibility, and the empirical work engages seriously with baselines and evaluation concerns. However, the central theorem is false as stated, and the proof chain contains a circular dependency between Proposition 1 and Theorem 1. Because the identifiability result and the algorithm's structural constraints both rest on Theorem 1, the paper's core theoretical contribution is not established. The empirical results, while suggestive, cannot compensate for the invalidity of the theory that motivates them.

major comments (4)
  1. [§4, Theorem 1 and Appendix A.2] The proof derives rank(J_f(x)) ≤ min(|Z|, |X|, |Y|) and then, after invoking Proposition 1, concludes rank(J_f(x)) = min(|Z|, |X|, |Y|) and subsequently equates this with |Z|. The final step is licensed only when |Z| ≤ min(|X|, |Y|). Condition 1 does not imply this bound: it explicitly allows a measured variable to have several latent parents (e.g., x12 in Figure 1), so the minimal d-separating set between two measured sets can be larger than either set. For example, take latent variables z1,...,z4 with measured children x1 (parents z1,z2), x2 (parents z3,z4), y1 (parents z1,z3), y2 (parents z2,z4), and add two pure children to each latent with unit path lengths to satisfy Condition 1. The smallest latent set d-separating X={x1,x2} from Y={y1,y2} has size 4, while J_f is a 2×2 matrix and hence has rank at most 2; for a generic faithful parameter choice its rank is 2, contradicting Eq. (3). This also breaks the claimed biconditional for r < |X|, |Y| when the minimal separator exceeds the Jacobian's maximal possible rank. Since Lemmas 1-3 and Theorem 3 all invoke Theorem 1, the identifiability theory is unsupported.
  2. [Appendix A.1 and Appendix A.2] The proof of Proposition 1 cites 'the proof of Theorem 1' for the decomposition J_h = J_f ∘ J_g, while the proof of Theorem 1 invokes Proposition 1 to conclude that the Jacobian attains its maximal rank almost everywhere. This is a direct circular dependency between the two results. In addition, Proposition 1's application of Lemma 5 requires |X| ≥ |Z| and |Y| ≥ |Z|; no such dimension condition is stated in Condition 1 or Condition 3, and the example in the previous comment shows that it can fail. The generic-rank claim is the bridge from the inequality rank(J_f) ≤ min(|Z|, |X|, |Y|) to the equality rank(J_f) = min(|Z|, |X|, |Y|), so this gap is load-bearing.
  3. [§4, Theorem 1 statement and Eq. (3)] The theorem's quantifier restricts attention to r < |X|, |Y|, but the formal display, Eq. (3) (and Eq. (11) in the proof), asserts an unqualified equality rank(J_f) = min_Z |Z|. The proof treats the unqualified equality as its target and then derives it by the invalid step identified above. As written, the formal equality is false for models where the minimal separator is larger than both |X| and |Y|. The restricted biconditional, even if it could be repaired under additional assumptions, is not sufficient for the subsequent lemmas and Theorem 3, which apply the equality to arbitrary measured sets without verifying the dimension bound.
  4. [§4, Theorem 3 and Appendix A.7] Theorem 3 assumes an oracle r(S,T) that outputs the minimum number of latent variables d-separating two measured sets, and the only route proposed in the paper for obtaining this oracle is the Jacobian rank criterion of Theorem 1. Since Theorem 1 is invalid, Theorem 3 does not establish identifiability from the observed distribution. Furthermore, the proof of Lemma 2 in Appendix A.5 asserts without derivation that any set d-separating S∪{c} from T must contain P and that r(S,T) = |P|; this presupposes the identification problem that the lemma is meant to solve. Thus the paper's central identifiability claim is not supported.
minor comments (4)
  1. [Appendix A.5, Lemma 2] In condition (1) of Lemma 2, the right-hand side reads '= P' where P is a set; it should presumably be '= |P|'.
  2. [Section 5.2, Eq. (8)] The constraint expression in Eq. (8) is malformed: it should read ∥Mi,:∥1 (∥Mi,: ⊙ ∏_{j≠i}(1−Mj,:)∥1 − 2) ≥ 0 for each i.
  3. [Figure 1 caption] The figure caption lists 'z8' twice in the sequence of latent variables; the second occurrence is likely a typo and should be a different latent variable.
  4. [Appendix B.1] There are typos, including 'signficant' and 'devation', and the explanation of the SHD-over-permutations evaluation is given twice and could be condensed.

Circularity Check

2 steps flagged · score 7.0 of 10

Central identifiability proof is mutually dependent (Proposition 1 is proved using Theorem 1's proof, while Theorem 1's proof invokes Proposition 1), and Theorem 1's claimed rank–d-separation equality is not licensed by its own proof.

  1. other [Appendix A.1 (Proof of Proposition 1) and Appendix A.2 (Proof of Theorem 1)]
    "Using the proof of Theorem 1, we know Jh(x) = Jf (g(x)) · Jg(x) where Jf (g(x)) ∈ R|Y|×|Z|, Jg ∈ R|Z|×|X|. ... Proposition 1 implies that the Jacobian achieves its maximal possible rank almost everywhere. Thus, using Condition 2, rank(Jf (x)) = min(|Z|, |X|, |Y|)."

    Proposition 1, which is the paper's justification of Generalized Faithfulness, is proved by explicitly borrowing the chain-rule factorization 'from the proof of Theorem 1'. Theorem 1's proof in turn uses Proposition 1 to upgrade its rank inequality to an equality. Neither result is proved independently: the maximal-rank fact on which Theorem 1's central equality rests is supplied by a proposition whose own proof imports the theorem it is used to prove. This mutual dependency makes the identifiability argument's core step circular.

  2. other [Theorem 1, Eq. (3), and proof in Appendix A.2]
    "rank(Jf (x)) ≤ min(|Z|, |X|, |Y|) ... rank(Jf (x)) = min(|Z|, |X|, |Y|) ... rank(Jf ) = min Z |Z| such that X ⊥⊥G Y|Z"

    The proof terminates at rank = min(|Z|, |X|, |Y|), but the theorem statement asserts rank = |Z|. The substitution min(|Z|, |X|, |Y|) = |Z| is valid only under |Z| ≤ min(|X|, |Y|), an assumption that is neither stated nor implied by Condition 1: Condition 1 explicitly allows a measured variable with several latent parents (x12 in Figure 1 has Pa(x12) = {z8, z9}), and then a minimal d-separating set can be larger than the maximum possible rank of the Jacobian. The claimed equality is therefore imposed rather than derived.

full rationale

The paper's headline theoretical claim, that identifiability follows from the observable rank of the Jacobian, rests on Theorem 1 and Proposition 1. Those two results exhibit a direct mutual proof dependency: the proof of Proposition 1 says 'Using the proof of Theorem 1...', while the proof of Theorem 1 says 'Proposition 1 implies...'. In addition, Theorem 1 as stated in Eq. (3) is stronger than the proof's conclusion: the proof only obtains rank = min(|Z|, |X|, |Y|), and the equality with |Z| needs an unstated bound |Z| ≤ min(|X|, |Y|) that Condition 1 does not provide. The later lemmas and Theorem 3 all rely on this rank–d-separation equivalence, so the central derivation is not self-contained. The algorithm itself is an independent engineering contribution with experiments against external baselines, and Theorem 3's oracle statement is a legitimate conditional combinatorial reduction rather than a circularity by itself. However, because the main theoretical result's proof contains the Proposition 1/Theorem 1 cycle and the unjustified min-to-|Z| reduction, the paper's central identifiability claim is not established by its own derivation. Score reflects partial circularity plus a load-bearing proof gap, not merely self-citation.

Assumptions & free parameters 5 free parameters · 4 assumptions · 0 invented entities

The central claim rests on structural conditions (pure children, equal path length, faithfulness, differentiability) and an unstated dimension bound on separating sets. No new physical entities are introduced. The algorithm adds hand-tuned hyperparameters that affect the learned graph.

free parameters (5)
  • lambda2 (L1 sparsity coefficient) = 1e-4 (synthetic), 0.03 (MNIST)
    Hand-chosen regularization weight; the learned graph depends on it.
  • lambda3 (pure-children constraint coefficient) = 1e-3 + epoch/100 (synthetic); exponential schedule 1e-3 to 10 (MNIST)
    Hand-scheduled penalty; the authors note the MNIST graph does not satisfy the constraint, so this parameter affects structure.
  • lambda1 (independence loss coefficient) = 10 (synthetic); not used for images
    Weight for the MINE-based independence penalty; omitted on image data.
  • Gumbel-softmax temperature = 1.0 (synthetic); 100 to 0.1 schedule (MNIST)
    Annealing schedule chosen by hand.
  • Maximum latent nodes per layer = 10, 20, 49 for MNIST; floor(|X|/2^i) for synthetic
    Model capacity choice; rows may be zeroed, but this bounds the search space.
assumptions (4)
  • domain assumption Each latent variable has at least two pure children and all its measured descendants are at equal path length (Condition 1)
    Required for identifiability; the paper admits it may not hold for learned MNIST graph.
  • domain assumption Generalized Faithfulness: rank Jacobian constraints that hold are entailed by the graph (Condition 2)
    Non-standard and not precisely defined; Proposition 1 attempts to justify it but the proof is circular.
  • domain assumption For every measured set X and latent set Z, p(z|x) depends on x through a differentiable sufficient statistic g(x) (Condition 3(ii))
    Strong assumption; the paper's own experiments violate it (non-differentiable LeakyReLU) yet claim identifiability.
  • ad hoc to paper The minimal d-separating latent set Z for any measured sets X,Y satisfies |Z| ≤ min(|X|,|Y|)
    Unstated in the paper; needed for Theorem 1's proof but false in general, e.g., singleton sets with a doubly-parented measured variable.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Differentiable Causal Discovery For Latent Hierarchical Causal Models." pith.science (2026). https://pith.science/paper/XARLJFNO

@misc{pith2026241119556,
  author       = {Pith},
  title        = {Pith review of: Differentiable Causal Discovery For Latent Hierarchical Causal Models},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/XARLJFNO}},
  note         = {Machine review of arXiv:2411.19556}
}
read the original abstract

Discovering causal structures with latent variables from observational data is a fundamental challenge in causal discovery. Existing methods often rely on constraint-based, iterative discrete searches, limiting their scalability to large numbers of variables. Moreover, these methods frequently assume linearity or invertibility, restricting their applicability to real-world scenarios. We present new theoretical results on the identifiability of nonlinear latent hierarchical causal models, relaxing previous assumptions in literature about the deterministic nature of latent variables and exogenous noise. Building on these insights, we develop a novel differentiable causal discovery algorithm that efficiently estimates the structure of such models. To the best of our knowledge, this is the first work to propose a differentiable causal discovery method for nonlinear latent hierarchical models. Our approach outperforms existing methods in both accuracy and scalability. We demonstrate its practical utility by learning interpretable hierarchical latent structures from high-dimensional image data and demonstrate its effectiveness on downstream tasks.

Figures

Figures reproduced from arXiv: 2411.19556 by the authors.

Figure 1
Figure 1. Example of a graph we consider. Note that we allow multiple paths between two nodes [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 2
Figure 2. Performance vs. Time for different causal discovery methods. Time is plotted on a [PITH_FULL_IMAGE:figures/full_fig_p009_2.png] view at source ↗
Figure 3
Figure 3. Figures for the Image experiments. (a) Latent causal graph for digit images (b) Visu [PITH_FULL_IMAGE:figures/full_fig_p010_3.png] view at source ↗
Figures from the paper (4 more)
Figure 4
Figure 4. Figure 4: Ground truth causal graphs for Synthetic Experiments. (a) and (b) are trees (only one path [PITH_FULL_IMAGE:figures/full_fig_p020_4.png]
Figure 5
Figure 5. Figure 5: Evolution of different loss components during training [PITH_FULL_IMAGE:figures/full_fig_p021_5.png]
Figure 6
Figure 6. Figure 6: Latent causal graph for the MNIST Dataset. [PITH_FULL_IMAGE:figures/full_fig_p024_6.png]
Figure 7
Figure 7. Figure 7: Two examples of causal structures which violate Condition 1 [PITH_FULL_IMAGE:figures/full_fig_p024_7.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

70 extracted references · 42 canonical work pages

  1. [1]

    Identification of partially observed linear causal models: Graphical conditions for the non-gaussian and heterogeneous cases

    Jeffrey Adams, Niels Hansen, and Kun Zhang. Identification of partially observed linear causal models: Graphical conditions for the non-gaussian and heterogeneous cases. Advances in Neural Information Processing Systems, 34: 0 22822--22833, 2021

    work page 2021

  2. [2]

    The decamfounder: nonlinear causal discovery in the presence of hidden variables

    Raj Agrawal, Chandler Squires, Neha Prasad, and Caroline Uhler. The decamfounder: nonlinear causal discovery in the presence of hidden variables. Journal of the Royal Statistical Society Series B: Statistical Methodology, 85 0 (5): 0 1639--1658, 2023

    work page 2023

  3. [3]

    Recursive causal structure learning in the presence of latent variables and selection bias

    Sina Akbari, Ehsan Mokhtarian, AmirEmad Ghassami, and Negar Kiyavash. Recursive causal structure learning in the presence of latent variables and selection bias. Advances in Neural Information Processing Systems, 34: 0 10119--10130, 2021

    work page 2021

  4. [4]

    Learning linear bayesian networks with latent variables

    Animashree Anandkumar, Daniel Hsu, Adel Javanmard, and Sham Kakade. Learning linear bayesian networks with latent variables. In International Conference on Machine Learning, pp.\ 249--257. PMLR, 2013

    work page 2013

  5. [5]

    Invariant risk minimization

    Martin Arjovsky, L \'e on Bottou, Ishaan Gulrajani, and David Lopez-Paz. Invariant risk minimization. arXiv preprint arXiv:1907.02893, 2019

    arXiv 1907

  6. [6]

    Mutual information neural estimation

    Mohamed Ishmael Belghazi, Aristide Baratin, Sai Rajeshwar, Sherjil Ozair, Yoshua Bengio, Aaron Courville, and Devon Hjelm. Mutual information neural estimation. In International conference on machine learning, pp.\ 531--540. PMLR, 2018

    work page 2018

  7. [7]

    Deconfounded score method: Scoring dags with dense unobserved confounding

    Alexis Bellot and Mihaela van der Schaar. Deconfounded score method: Scoring dags with dense unobserved confounding. arXiv preprint arXiv:2103.15106, 2021

    arXiv 2021

  8. [8]

    Differentiable causal discovery under unmeasured confounding

    Rohit Bhattacharya, Tushar Nagarajan, Daniel Malinsky, and Ilya Shpitser. Differentiable causal discovery under unmeasured confounding. In International Conference on Artificial Intelligence and Statistics, pp.\ 2314--2322. PMLR, 2021

    2021

Show all 70 references
  1. [9]

    Weakly supervised causal representation learning

    Johann Brehmer, Pim De Haan, Phillip Lippe, and Taco S Cohen. Weakly supervised causal representation learning. Advances in Neural Information Processing Systems, 35: 0 38319--38331, 2022

    2022

  2. [10]

    Differentiable causal discovery from interventional data

    Philippe Brouillard, S \'e bastien Lachapelle, Alexandre Lacoste, Simon Lacoste-Julien, and Alexandre Drouin. Differentiable causal discovery from interventional data. Advances in Neural Information Processing Systems, 33: 0 21865--21877, 2020

    2020

  3. [11]

    Identification of linear latent variable model with arbitrary distribution

    Zhengming Chen, Feng Xie, Jie Qiao, Zhifeng Hao, Kun Zhang, and Ruichu Cai. Identification of linear latent variable model with arbitrary distribution. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 36, pp.\ 6350--6357, 2022

    2022

  4. [12]

    Optimal structure identification with greedy search

    David Maxwell Chickering. Optimal structure identification with greedy search. Journal of machine learning research, 3 0 (Nov): 0 507--554, 2002

    2002

  5. [13]

    Large-sample learning of bayesian networks is np-hard

    Max Chickering, David Heckerman, and Chris Meek. Large-sample learning of bayesian networks is np-hard. Journal of Machine Learning Research, 5: 0 1287--1330, 2004

    2004

  6. [14]

    Learning latent tree graphical models

    Myung Jin Choi, Vincent YF Tan, Animashree Anandkumar, and Alan S Willsky. Learning latent tree graphical models. The Journal of Machine Learning Research, 12: 0 1771--1812, 2011

    2011

  7. [15]

    Learning sparse causal models is not np-hard

    Tom Claassen, Joris Mooij, and Tom Heskes. Learning sparse causal models is not np-hard. arXiv preprint arXiv:1309.6824, 2013

    2013 arXiv

  8. [16]

    Learning high-dimensional directed acyclic graphs with latent and selection variables

    Diego Colombo, Marloes H Maathuis, Markus Kalisch, and Thomas S Richardson. Learning high-dimensional directed acyclic graphs with latent and selection variables. The Annals of Statistics, pp.\ 294--321, 2012

    2012

  9. [17]

    Learning the causal structure of copula models with latent variables

    Ruifei Cui, Perry Groot, Moritz Schauer, and Tom Heskes. Learning the causal structure of copula models with latent variables. 2018

    2018

  10. [18]

    A versatile causal discovery framework to allow causally-related hidden variables

    Xinshuai Dong, Biwei Huang, Ignavier Ng, Xiangchen Song, Yujia Zheng, Songyao Jin, Roberto Legaspi, Peter Spirtes, and Kun Zhang. A versatile causal discovery framework to allow causally-related hidden variables. arXiv preprint arXiv:2312.11001, 2023

    2023 arXiv

  11. [19]

    Asymptotic evaluation of certain markov process expectations for large time

    Monroe D Donsker and SR Srinivasa Varadhan. Asymptotic evaluation of certain markov process expectations for large time. iv. Communications on pure and applied mathematics, 36 0 (2): 0 183--212, 1983

    1983

  12. [20]

    Marginal likelihood and model selection for gaussian latent tree and forest models

    Mathias Drton, Shaowei Lin, Luca Weihs, and Piotr Zwiernik. Marginal likelihood and model selection for gaussian latent tree and forest models. 2017

    2017

  13. [21]

    Unsupervised learning of transcriptional regulatory networks via latent tree graphical models

    Anthony Gitter, Furong Huang, Ragupathyraj Valluvan, Ernest Fraenkel, and Animashree Anandkumar. Unsupervised learning of transcriptional regulatory networks via latent tree graphical models. arXiv preprint arXiv:1609.06335, 2016

    2016 arXiv

  14. [22]

    Bayesian pyramids: Identifiable multilayer discrete latent structure models for discrete data

    Yuqi Gu and David B Dunson. Bayesian pyramids: Identifiable multilayer discrete latent structure models for discrete data. Journal of the Royal Statistical Society Series B: Statistical Methodology, 85 0 (2): 0 399--426, 2023

    2023

  15. [23]

    Variational autoencoders with jointly optimized latent dependency structure

    Jiawei He, Yu Gong, Joseph Marino, Greg Mori, and Andreas Lehrmann. Variational autoencoders with jointly optimized latent dependency structure. In International conference on learning representations, 2018

    2018

  16. [24]

    beta-vae: Learning basic visual concepts with a constrained variational framework

    Irina Higgins, Loic Matthey, Arka Pal, Christopher P Burgess, Xavier Glorot, Matthew M Botvinick, Shakir Mohamed, and Alexander Lerchner. beta-vae: Learning basic visual concepts with a constrained variational framework. ICLR (Poster), 3, 2017 a

    2017

  17. [25]

    Scan: Learning hierarchical compositional visual concepts

    Irina Higgins, Nicolas Sonnerat, Loic Matthey, Arka Pal, Christopher P Burgess, Matko Bosnjak, Murray Shanahan, Matthew Botvinick, Demis Hassabis, and Alexander Lerchner. Scan: Learning hierarchical compositional visual concepts. arXiv preprint arXiv:1707.03389, 2017 b

    2017 arXiv

  18. [26]

    Latent hierarchical causal structure discovery with rank constraints

    Biwei Huang, Charles Jia Han Low, Feng Xie, Clark Glymour, and Kun Zhang. Latent hierarchical causal structure discovery with rank constraints. Advances in neural information processing systems, 35: 0 5549--5561, 2022

    2022

  19. [27]

    Nonlinear ica using auxiliary variables and generalized contrastive learning

    Aapo Hyvarinen, Hiroaki Sasaki, and Richard Turner. Nonlinear ica using auxiliary variables and generalized contrastive learning. In The 22nd International Conference on Artificial Intelligence and Statistics, pp.\ 859--868. PMLR, 2019

    2019

  20. [28]

    Categorical reparameterization with gumbel-softmax

    Eric Jang, Shixiang Gu, and Ben Poole. Categorical reparameterization with gumbel-softmax. arXiv preprint arXiv:1611.01144, 2016

    2016 arXiv

  21. [29]

    Auto-encoding variational bayes

    Diederik P Kingma. Auto-encoding variational bayes. arXiv preprint arXiv:1312.6114, 2013

    2013 arXiv

  22. [30]

    Learning latent causal graphs via mixture oracles

    Bohdan Kivva, Goutham Rajendran, Pradeep Ravikumar, and Bryon Aragam. Learning latent causal graphs via mixture oracles. Advances in Neural Information Processing Systems, 34: 0 18087--18101, 2021

    2021

  23. [31]

    Identification of nonlinear latent hierarchical models

    Lingjing Kong, Biwei Huang, Feng Xie, Eric Xing, Yuejie Chi, and Kun Zhang. Identification of nonlinear latent hierarchical models. Advances in Neural Information Processing Systems, 36: 0 2010--2032, 2023

    2010

  24. [32]

    Learning discrete concepts in latent hierarchical models

    Lingjing Kong, Guangyi Chen, Biwei Huang, Eric P Xing, Yuejie Chi, and Kun Zhang. Learning discrete concepts in latent hierarchical models. arXiv preprint arXiv:2406.00519, 2024

    2024 arXiv

  25. [33]

    Causal clustering for 1-factor measurement models

    Erich Kummerfeld and Joseph Ramsey. Causal clustering for 1-factor measurement models. In Proceedings of the 22nd ACM SIGKDD international conference on knowledge discovery and data mining, pp.\ 1655--1664, 2016

    2016

  26. [34]

    Mnist handwritten digit database

    Yann LeCun, Corinna Cortes, and CJ Burges. Mnist handwritten digit database. ATT Labs [Online]. Available: http://yann.lecun.com/exdb/mnist, 2, 2010

    2010

  27. [35]

    Causal discovery from observational and interventional data across multiple environments

    Adam Li, Amin Jaber, and Elias Bareinboim. Causal discovery from observational and interventional data across multiple environments. Advances in Neural Information Processing Systems, 36: 0 16942--16956, 2023

    2023

  28. [36]

    Deep learning face attributes in the wild

    Ziwei Liu, Ping Luo, Xiaogang Wang, and Xiaoou Tang. Deep learning face attributes in the wild. In Proceedings of the IEEE international conference on computer vision, pp.\ 3730--3738, 2015

    2015

  29. [37]

    Scalable differentiable causal discovery in the presence of latent confounders with skeleton posterior (extended version)

    Pingchuan Ma, Rui Ding, Qiang Fu, Jiaru Zhang, Shuai Wang, Shi Han, and Dongmei Zhang. Scalable differentiable causal discovery in the presence of latent confounders with skeleton posterior (extended version). arXiv preprint arXiv:2406.10537, 2024

    2024 arXiv

  30. [38]

    Stable differentiable causal discovery

    Achille Nazaret, Justin Hong, Elham Azizi, and David Blei. Stable differentiable causal discovery. arXiv preprint arXiv:2311.10263, 2023

    2023 arXiv

  31. [39]

    Masked gradient-based causal structure learning

    Ignavier Ng, Shengyu Zhu, Zhuangyan Fang, Haoyang Li, Zhitang Chen, and Jun Wang. Masked gradient-based causal structure learning. In Proceedings of the 2022 SIAM International Conference on Data Mining (SDM), pp.\ 424--432. SIAM, 2022

    2022

  32. [40]

    Structure learning with continuous optimization: A sober look and beyond

    Ignavier Ng, Biwei Huang, and Kun Zhang. Structure learning with continuous optimization: A sober look and beyond. In Causal Learning and Reasoning, pp.\ 71--105. PMLR, 2024

    2024

  33. [41]

    Comprehensive review and empirical evaluation of causal discovery algorithms for numerical data

    Wenjin Niu, Zijun Gao, Liyan Song, and Lingbo Li. Comprehensive review and empirical evaluation of causal discovery algorithms for numerical data. arXiv preprint arXiv:2407.13054, 2024

    2024 arXiv

  34. [42]

    Causes of severe pneumonia requiring hospital admission in children without hiv infection from africa and asia: the perch multi-country case-control study

    Katherine L O'Brien, Henry C Baggett, W Abdullah Brooks, Daniel R Feikin, Laura L Hammitt, Melissa M Higdon, Stephen RC Howie, Maria Deloria Knoll, Karen L Kotloff, Orin S Levine, et al. Causes of severe pneumonia requiring hospital admission in children without hiv infection ...

    2019

  35. [43]

    Probabilistic reasoning in intelligent systems; network of plausible inference

    J Pearl. Probabilistic reasoning in intelligent systems; network of plausible inference. Morgan Kaufmann, 1988, 1988

    1988

  36. [44]

    Models, reasoning and inference

    Judea Pearl et al. Models, reasoning and inference. Cambridge, UK: CambridgeUniversityPress, 19 0 (2): 0 3, 2000

    2000

  37. [45]

    Beware of the simulated dag! causal discovery benchmarks may be easy to game

    Alexander Reisach, Christof Seiler, and Sebastian Weichwald. Beware of the simulated dag! causal discovery benchmarks may be easy to game. Advances in Neural Information Processing Systems, 34: 0 27772--27784, 2021

    2021

  38. [46]

    Toward causal representation learning

    Bernhard Sch \"o lkopf, Francesco Locatello, Stefan Bauer, Nan Rosemary Ke, Nal Kalchbrenner, Anirudh Goyal, and Yoshua Bengio. Toward causal representation learning. Proceedings of the IEEE, 109 0 (5): 0 612--634, 2021

    2021

  39. [47]

    Learning large dags is harder than you think: Many losses are minimal for the wrong dag

    Jonas Seng, Matej Ze c evi \'c , Devendra Singh Dhami, and Kristian Kersting. Learning large dags is harder than you think: Many losses are minimal for the wrong dag. In The Twelfth International Conference on Learning Representations, 2024

    2024

  40. [48]

    Nodags-flow: Nonlinear cyclic causal structure learning

    Muralikrishnna G Sethuraman, Romain Lopez, Rahul Mohan, Faramarz Fekri, Tommaso Biancalani, and Jan-Christian H \"u tter. Nodags-flow: Nonlinear cyclic causal structure learning. In International Conference on Artificial Intelligence and Statistics, pp.\ 6371--6387. PMLR, 2023

    2023

  41. [49]

    Estimation of linear non-gaussian acyclic models for latent factors

    Shohei Shimizu, Patrik O Hoyer, and Aapo Hyv \"a rinen. Estimation of linear non-gaussian acyclic models for latent factors. Neurocomputing, 72 0 (7-9): 0 2024--2027, 2009

    2024

  42. [50]

    Learning the structure of linear latent variable models

    Ricardo Silva, Richard Scheines, Clark Glymour, Peter Spirtes, and David Maxwell Chickering. Learning the structure of linear latent variable models. Journal of Machine Learning Research, 7 0 (2), 2006

    2006

  43. [51]

    Introduction to causal inference

    Peter Spirtes. Introduction to causal inference. Journal of Machine Learning Research, 11 0 (5), 2010

    2010

  44. [52]

    Causation, prediction, and search

    Peter Spirtes, Clark Glymour, and Richard Scheines. Causation, prediction, and search. MIT press, 2001

    2001

  45. [53]

    Unpaired multi-domain causal representation learning

    Nils Sturma, Chandler Squires, Mathias Drton, and Caroline Uhler. Unpaired multi-domain causal representation learning. Advances in Neural Information Processing Systems, 36, 2024

    2024

  46. [54]

    Learning latent structural causal models

    Jithendaraa Subramanian, Yashas Annadani, Ivaxi Sheth, Nan Rosemary Ke, Tristan Deleu, Stefan Bauer, Derek Nowrouzezahrai, and Samira Ebrahimi Kahou. Learning latent structural causal models. arXiv preprint arXiv:2210.13583, 2022

    arXiv

  47. [55]

    Nvae: A deep hierarchical variational autoencoder

    Arash Vahdat and Jan Kautz. Nvae: A deep hierarchical variational autoencoder. Advances in neural information processing systems, 33: 0 19667--19679, 2020

    2020

  48. [56]

    Hierarchical causal models

    Eli N Weinstein and David M Blei. Hierarchical causal models. arXiv preprint arXiv:2401.05330, 2024

    2024 arXiv

  49. [57]

    Generalized independent noise condition for estimating latent variable causal graphs

    Feng Xie, Ruichu Cai, Biwei Huang, Clark Glymour, Zhifeng Hao, and Kun Zhang. Generalized independent noise condition for estimating latent variable causal graphs. Advances in neural information processing systems, 33: 0 14891--14902, 2020

    2020

  50. [58]

    Identification of linear non-gaussian latent hierarchical structure

    Feng Xie, Biwei Huang, Zhengming Chen, Yangbo He, Zhi Geng, and Kun Zhang. Identification of linear non-gaussian latent hierarchical structure. In International Conference on Machine Learning, pp.\ 24370--24387. PMLR, 2022

    2022

  51. [59]

    Causalvae: Disentangled representation learning via neural structural causal models

    Mengyue Yang, Furui Liu, Zhitang Chen, Xinwei Shen, Jianye Hao, and Jun Wang. Causalvae: Disentangled representation learning via neural structural causal models. In Proceedings of the IEEE/CVF conference on computer vision and pattern recognition, pp.\ 9593--9602, 2021

    2021

  52. [60]

    Dag-gnn: Dag structure learning with graph neural networks

    Yue Yu, Jie Chen, Tian Gao, and Mo Yu. Dag-gnn: Dag structure learning with graph neural networks. In International conference on machine learning, pp.\ 7154--7163. PMLR, 2019

    2019

  53. [61]

    Causal discovery with multi-domain lingam for latent factors

    Yan Zeng, Shohei Shimizu, Ruichu Cai, Feng Xie, Michio Yamamoto, and Zhifeng Hao. Causal discovery with multi-domain lingam for latent factors. In Causal Analysis Workshop Series, pp.\ 1--4. PMLR, 2021

    2021

  54. [62]

    On the completeness of orientation rules for causal discovery in the presence of latent confounders and selection bias

    Jiji Zhang. On the completeness of orientation rules for causal discovery in the presence of latent confounders and selection bias. Artificial Intelligence, 172 0 (16-17): 0 1873--1896, 2008

    2008

  55. [63]

    D-vae: A variational autoencoder for directed acyclic graphs

    Muhan Zhang, Shali Jiang, Zhicheng Cui, Roman Garnett, and Yixin Chen. D-vae: A variational autoencoder for directed acyclic graphs. Advances in neural information processing systems, 32, 2019

    2019

  56. [64]

    Dags with no tears: Continuous optimization for structure learning

    Xun Zheng, Bryon Aragam, Pradeep K Ravikumar, and Eric P Xing. Dags with no tears: Continuous optimization for structure learning. Advances in neural information processing systems, 31, 2018

    2018

  57. [65]

    Learning sparse nonparametric dags

    Xun Zheng, Chen Dan, Bryon Aragam, Pradeep Ravikumar, and Eric Xing. Learning sparse nonparametric dags. In International Conference on Artificial Intelligence and Statistics, pp.\ 3414--3425. Pmlr, 2020

    2020

  58. [66]

    On the identifiability of nonlinear ica: Sparsity and beyond

    Yujia Zheng, Ignavier Ng, and Kun Zhang. On the identifiability of nonlinear ica: Sparsity and beyond. Advances in neural information processing systems, 35: 0 16411--16422, 2022

    2022

  59. [67]

    write newline

    " write newline "" before.all 'output.state := FUNCTION n.dashify 't := "" t empty not t #1 #1 substring "-" = t #1 #2 substring "--" = not "--" * t #2 global.max substring 't := t #1 #1 substring "-" = "-" * t #2 global.max substring 't := while if t #1 #1 substring * t #2 gl...

  60. [68]

    @esa (Ref

    \@ifxundefined[1] #1\@undefined \@firstoftwo \@secondoftwo \@ifnum[1] #1 \@firstoftwo \@secondoftwo \@ifx[1] #1 \@firstoftwo \@secondoftwo [2] @ #1 \@temptokena #2 #1 @ \@temptokena \@ifclassloaded agu2001 natbib The agu2001 class already includes natbib coding, so you should ...

  61. [69]

    \@lbibitem[] @bibitem@first@sw\@secondoftwo \@lbibitem[#1]#2 \@extra@b@citeb \@ifundefined br@#2\@extra@b@citeb \@namedef br@#2 \@nameuse br@#2\@extra@b@citeb \@ifundefined b@#2\@extra@b@citeb @num @parse #2 @tmp #1 NAT@b@open@#2 NAT@b@shut@#2 \@ifnum @merge>\@ne @bibitem@firs...

  62. [70]

    silva2006learning and kummerfeld2016causal utilize tetrad conditions---the rank of each 2 2 sub-covariance matrix---to discover latent variables

    @open @close @open @close and [1] URL: #1 \@ifundefined chapter * \@mkboth \@ifxundefined @sectionbib * \@mkboth * \@mkboth\@gobbletwo \@ifclassloaded amsart * \@ifclassloaded amsbook * \@ifxundefined @heading @heading NAT@ctr thebibliography [1] @ \@biblabel @NAT@ctr \@bibset...

Pith tools

Reviewed August 12, 2026 · model on record in the stance chip above.