Semi-Implicit Graph Variational Auto-Encoders

Arman Hasanzadeh; Ehsan Hajiramezanali; Krishna R. Narayanan; Mingyuan Zhou; Nick Duffield; Xiaoning Qian

REVIEW 3 major objections 4 minor 42 references

Semi-Implicit Graph Variational Auto-Encoders

T0 review · 3 major / 4 minor · reviewed 2026-08-14 · deepseek-v4-flash

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

Pith's one-line read Injecting random noise at every layer of a graph auto-encoder's graph-neural-network encoder, and pairing it with a Bernoulli-Poisson decoder, yields implicit node posteriors that consistently beat link-prediction baselines.

desk verdict The empirical case is solid; the variational theory is more borrowed than established in this setting, but the architecture and ablations carry the paper. read the letter →

arxiv 1908.07078 v4 pith:TYAOGLWG submitted 2019-08-19 cs.LG stat.ML

Arman Hasanzadeh , Ehsan Hajiramezanali , Nick Duffield , Krishna R. Narayanan , Mingyuan Zhou , Xiaoning Qian This is my paper · ORCID

classification cs.LGstat.ML

keywords semi-implicitvariationalinferencegraphauto-encoderneuralnetworkslinkpredictionBernoulli-Poissondecodernodeembeddingimplicitposteriorgeneration

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 establish that the two main weaknesses of variational graph auto-encoders are fixable in a single architecture. The first weakness is the Gaussian assumption on the node embedding posterior, which cannot represent heavy-tailed, skewed, or multimodal distributions. The second is the inner-product decoder, which cannot reproduce the sparsity of real-world graphs. The proposed model, SIG-VAE, injects random noise at every graph-neural-network layer so that each node's latent distribution is formed by mixing the distributions of its neighbors, and it pairs this hierarchical stochastic encoder with a Bernoulli-Poisson link decoder. The paper reports consistent link-prediction improvements over existing methods on eight datasets, along with more interpretable embeddings and generated graphs whose density and clustering coefficients closely match the training graph.

What carries the argument

The load-bearing construction is a stochastic hierarchical encoder: at each layer $u$, the node attributes are concatenated with random noise $\epsilon_u$ and passed through a graph neural network, so the output mean and covariance are random variables; after marginalizing out the noise, the posterior is no longer Gaussian. Inference is carried out with a surrogate ELBO derived from semi-implicit variational inference, which lower-bounds the true ELBO by moving the KL divergence inside the expectation over the mixing distribution, with a regularization term that prevents collapse to a point mass. The decoder is a Bernoulli-Poisson link, $p(A_{ij}=1 \mid z_i,z_j,R)=1-e^{-\exp(\sum_k r_k z_{ik} z_{jk})}$, where $R$ is a diagonal matrix of learned positive coefficients; this makes edges sparse and lets the model generate graphs with realistic density.

What would settle it

Train SIG-VAE on a synthetic graph with planted overlapping communities where the true posterior is known to be a mixture, then inspect whether the inferred per-node distributions match the planted mixture weights; if the surrogate objective is loose or biased, the posterior will collapse to a single mode per node. A simpler observable check is to ablate the SIVI regularizer: if the reported gains come from that regularizer rather than from the hierarchical construction, AUC and latent variance should drop sharply on the Power graph.

Watch

Extended reading notes

Core claim

SIG-VAE claims that a hierarchical semi-implicit variational encoder, in which the parameters of the node-embedding distribution are themselves random variables produced by graph-convolution layers, can faithfully approximate implicit posteriors over graph embeddings. Because the marginal posterior is intractable, the paper optimizes a surrogate evidence lower bound borrowed from semi-implicit variational inference, adding a nonnegative regularizer that makes the bound asymptotically exact. On the generative side, the decoder replaces the inner product with a Bernoulli-Poisson link function, modeling each edge as the indicator of a latent Poisson count. The paper demonstrates on a Swiss-roll graph that inferred node distributions become multimodal near community boundaries, and that on link prediction SIG-VAE achieves higher AUC and average precision than VGAE and other baselines, for example improving AUC on the Power graph from 71.20 to 96.23.

Load-bearing premise

The load-bearing premise is that the surrogate ELBO used for training, with the nonnegative regularizer taken from semi-implicit variational inference, is tight enough that optimizing it recovers the true implicit posterior, and the paper does not re-verify this property for its multi-layer stochastic graph encoder.

Editorial extensions

If this is right

Link prediction improves consistently across both attributed and unattributed graphs; on the Power graph, AUC rises from 71.20 with VGAE to 96.23 with SIG-VAE.
Nodes that sit between communities receive multimodal latent posteriors, so the model's uncertainty encodes structural ambiguity rather than a fixed Gaussian spread.
The generative model produces graphs with global statistics close to the training graph: for Cora, generated density is 0.00147 versus 0.00143 and average clustering coefficient is 0.25 versus 0.24.
Replacing only the decoder with the Bernoulli-Poisson link, or only the posterior with a normalizing flow, yields smaller gains than the full model, indicating that the hierarchical noise sharing rather than any single component drives the improvement.
The same hyperparameters work across all datasets, suggesting the method is not tuned to a particular graph size or to the presence of node attributes.

Reading between the lines

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

If the surrogate ELBO is as tight as claimed, the same noise-injection recipe could be transferred to other graph generative settings, such as temporal or heterogeneous graphs, where the posterior over latent structure is likely multimodal.
The two-stage training used for unattributed graphs hints at a general strategy: learn a coarse high-dimensional embedding first, then refine it with high noise; this curriculum could be tested as a standalone improvement on attributed graphs.
A direct test of posterior fidelity would be to compare the model's predictive confidence on link prediction against empirical accuracy; faithful implicit posteriors should give well-calibrated uncertainty, which would be valuable for high-stakes link prediction such as drug-target interaction.
The sparse-graph generation result suggests the Bernoulli-Poisson decoder could serve as a prior for molecule or knowledge-graph generation, where edge sparsity and community structure are the dominant statistics.

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.

Desk Editor's Note

read the letter

First, what you should know: the paper is an empirical win. SIG-VAE, which injects noise into every GNN layer and uses a Bernoulli-Poisson decoder, beats VGAE and a bunch of baselines on eight link prediction datasets, sometimes by a large margin (Power AUC 71 to 96). The ablations are the right ones: they compare against Naive SIG-VAE (SIVI without noise propagation) and NF-VGAE, so you can see that the neighborhood sharing is doing real work. The code is out, and the experiments are described in enough detail to reproduce. That is the core of the paper and it holds up.

The new idea is the hierarchical stochastic encoder: by making the inputs to each GNN layer random, the distribution parameters of neighboring nodes mix before the decoding step. That is a genuine extension of SIVI to graphs, not a trivial combination. The Bernoulli-Poisson decoder is prior work from one of the authors, but pairing it with this encoder is new and yields the sparse-graph gains.

Where I'd push back: the surrogate ELBO. The paper defines the ELBO in Eq. (8), then says the surrogate is derived in the supplementary and cites SIVI for asymptotic exactness. The regularizer isn't written down in the main text. For a NeurIPS paper, that's a minor presentation issue, not a fatal one—SIVI's theory is established, and the supplementary is the right place for the algebra. But if I were refereeing, I'd ask for the graph-specific derivation and a check that the regularizer doesn't degenerate here. The concern that the reported gains could be an artifact of optimizing a loose bound is real in principle, but the ablations against Naive SIG-VAE, which uses the same SIVI objective, isolate the architecture as the cause. So I don't think this sinks the paper.

Other soft spots: Table 2 doesn't show error bars (they're in the supplementary), and the two-stage learning for attribute-free graphs is a practical hack, but it's applied consistently to baselines. The abstract overclaims 'significantly outperforms state-of-the-art'—true for link prediction, but node classification results are only comparable, not superior.

Who's this for? Anyone working on graph representation learning or variational inference for graphs. The paper deserves a serious referee. I'd accept it with requests for more attentive treatment of the surrogate and error bars in the main text.

Referee Report

3 major / 4 minor

Summary. The paper proposes SIG-VAE, a semi-implicit graph variational auto-encoder that combines a hierarchical stochastic encoder with a Bernoulli-Poisson link decoder. The encoder injects random noise at multiple GNN layers, so the parameters of the per-node Gaussian conditional are themselves random variables; marginalizing them yields an implicit and potentially non-Gaussian posterior. The decoder models each edge as a thresholded Poisson variable, which is intended to better capture sparse real-world graphs. The authors compare SIG-VAE against VGAE, a naive SIVI-VGAE, an NF-VGAE, and several other baselines on link prediction, graph generation, node classification, and latent-representation interpretability, reporting strong link-prediction results on eight datasets and an ablation that isolates the effect of the hierarchical noise injection.

Significance. If the reported results are correct, SIG-VAE constitutes a practical improvement over VGAE and related variational graph embedding methods, with consistent gains on attribute-free and attributed graphs and a decoder that generates sparse graphs with realistic statistics. The paper's strengths include an extensive comparison against external baselines, ablations that separate the hierarchical inference scheme from the Bernoulli-Poisson decoder, and a public implementation. The main weakness is that the optimization objective's theoretical justification is cited from prior work rather than derived or verified for the specific hierarchical stochastic GNN encoder, and this is load-bearing for the empirical claims.

major comments (3)

[Section 4.1, Eq. (8)] The surrogate lower bound L = -E_psi KL(q(Z|psi)||p(Z)) + E_psi E_Z log p(A|Z) is stated to be a lower bound on the ELBO and asymptotically exact because of SIVI's regularization, but the regularized objective is never written down and no argument is given that the SIVI guarantee, proven for a generic conditional q(Z|psi), carries over to the hierarchical stochastic encoder in Eqs. (5)-(6) where psi=(mu,Sigma) depends on noise injected at every GNN layer. Since every reported result is obtained by optimizing this surrogate, the missing derivation and the unspecified regularizer are load-bearing; the pointer to the supplementary material is not sufficient for the paper as submitted.
[Table 2] The claim that SIG-VAE shows 'consistent superior performance' is not supported on the USAir graph: SIG-VAE's AUC and AP (94.52 and 94.95) are lower than those of SIG-VAE(IP) (97.56 and 97.50) and its AUC is below SEAL's 97.09. The paper notes that the Bernoulli-Poisson decoder is better suited to sparser graphs, but the abstract and Section 5.2 should be qualified so that the blanket statement about significantly outperforming state-of-the-art methods reflects the actual pattern of results.
[Section 5.2, attribute-free datasets] The two-stage learning procedure for graphs without node attributes (128-dimensional embedding with 5-dimensional Bernoulli noise, then 16-dimensional embedding with 64-dimensional noise) is described only briefly and without stating how the hyperparameters of the two stages were selected. Since all attribute-free link-prediction results in Table 2 depend on this protocol, the paper should report the selection criterion or demonstrate that the conclusions are robust to reasonable variations in these choices.

minor comments (4)

[Section 5.2] There are typos in the text: 'attribbutes' should be 'attributes' and 'Berhoulli-Poisson' should be 'Bernoulli-Poisson'.
[Section 4, paragraph after Eq. (5)] The sentence 'Note that (5) is different from the NF-VAE construction (3)' appears to cite the wrong equation: Eq. (3) defines SIVI-VGAE, while NF-VGAE is defined in Eq. (4).
[Table 2] Table 2 omits standard deviations in the main text, despite the significance claims relying on small margins in several rows (e.g., Yeast AUC 98.32 vs. 98.11); the standard deviations should be reported in the main table or the main text should at least summarize the variability range.
[Equations (5) and (7)] The symbol L is used both for the number of stochastic layers in the encoder and for the latent dimension in the Bernoulli-Poisson decoder; please disambiguate the notation, for example by using a separate symbol for the number of layers.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the central link-prediction claims are empirical and tested against external baselines.

full rationale

The model's derivation chain is not circular. The hierarchical encoder in Eqs. (5)-(6) is a new architectural proposal, while the variational lower bound in Section 4.1 applies the Jensen-based SIVI bound from Yin and Zhou (2018), which is prior independent published work rather than a restatement of this paper's conclusions. The Bernoulli-Poisson decoder is likewise adopted from Zhou (2015), a separate external model. The main empirical claims are evaluated against held-out test edges and compared with external baselines (VGAE, GAE, SEAL, S-VGAE, etc.), so the reported AUC/AP values are not fitted quantities renamed as predictions. The self-citations to the SIVI and Bernoulli-Poisson papers are real, but they are not circular because the cited results do not assume or depend on the present paper's target result. One completeness concern is that the regularized surrogate ELBO derivation is deferred to the supplementary material, which weakens verification of the optimization objective, but this is a missing-support/correctness issue rather than a circularity. No load-bearing step reduces by construction to the paper's own inputs.

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

The model rests on the SIVI surrogate ELBO and the Bernoulli-Poisson decoder, both imported from co-authored prior work. The free parameters listed are architectural choices (latent dimensions and noise sizes) and a threshold used in the graph generation demonstration; they are not fitted to force the reported link prediction results.

free parameters (4)

Latent dimension d = 16 (attribute graphs); 128 then 16 (attribute-free two-stage)
Chosen by hand; the paper sets d=16 in Section 5.2 and uses 128/16 in the two-stage procedure for datasets without attributes.
Noise dimension for epsilon_u = 5 (first stage) and 64 (second stage) on attribute-free datasets
Injected Bernoulli noise dimensions at each GNN layer, selected empirically rather than by a search or theory.
Number of stochastic layers L = 2 (for Cora classification setup, hidden [32,16])
Architecture depth chosen by the authors; not swept in the experiments.
Graph generation threshold for r_k = 0.01
In Section 5.3, inferred r_k values smaller than 0.01 are set to 0 to generate matching graph statistics; this is a manual post-hoc adjustment for the generation demo.

assumptions (4)

standard math Jensen's inequality: E_psi KL(q(Z|psi)||p(Z)) >= KL(E_psi[q(Z|psi)]||p(Z))
Used in Eq. (2) to define the surrogate ELBO lower bound; without this the semi-implicit objective is intractable.
domain assumption SIVI's surrogate ELBO is asymptotically exact under a nonnegative regularization term
Invoked in Section 4.1 to justify optimizing the surrogate; asserted from Yin and Zhou 2018 and not re-derived for the graph setting.
standard math The reparameterization trick applies to the Gaussian q(Z|psi) and the Bernoulli noise epsilon_u
Needed to backpropagate through random samples in Eqs. (5)-(6).
domain assumption The Bernoulli-Poisson link is a suitable generative model for sparse graphs
Adopted from Zhou 2015 in Eq. (7); the paper uses it to model edge probabilities as 1-exp(-exp(sum r_k z_ik z_jk)).

how reviews work

0 comments

Cite this review

Pith. "Pith review of Semi-Implicit Graph Variational Auto-Encoders." pith.science (2026). https://pith.science/paper/TYAOGLWG

@misc{pith2026190807078,
  author       = {Pith},
  title        = {Pith review of: Semi-Implicit Graph Variational Auto-Encoders},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/TYAOGLWG}},
  note         = {Machine review of arXiv:1908.07078}
}

read the original abstract

Semi-implicit graph variational auto-encoder (SIG-VAE) is proposed to expand the flexibility of variational graph auto-encoders (VGAE) to model graph data. SIG-VAE employs a hierarchical variational framework to enable neighboring node sharing for better generative modeling of graph dependency structure, together with a Bernoulli-Poisson link decoder. Not only does this hierarchical construction provide a more flexible generative graph model to better capture real-world graph properties, but also does SIG-VAE naturally lead to semi-implicit hierarchical variational inference that allows faithful modeling of implicit posteriors of given graph data, which may exhibit heavy tails, multiple modes, skewness, and rich dependency structures. Compared to VGAE, the derived graph latent representations by SIG-VAE are more interpretable, due to more expressive generative model and more faithful inference enabled by the flexible semi-implicit construction. Extensive experiments with a variety of graph data show that SIG-VAE significantly outperforms state-of-the-art methods on several different graph analytic tasks.

Figures

Figures reproduced from arXiv: 1908.07078 by the authors.

**Figure 1.** SIG-VAE diffuses the distributions of the neighboring nodes, which is more informative than sharing deterministic features, to infer each node’s latent distribution. In SIG-VAE, when inferring each node’s latent posterior, we incorporate the distributions of the neighboring nodes, better capturing graph dependency structure than sharing deterministic features from GNNs. More specifically, the input to our model at… view at source ↗

**Figure 2.** Swiss roll graph (left) and its latent representation using SIG-VAE (middle) and VGAE (right). The latent representations (middle and right) are heat maps in R 3 . We expect that the embedding of the Swiss roll graph with inner-product decoder to be a curved plane in R 3 , which is clearly captured better by SIG-VAE. 2.5 2.0 1.5 1.0 0.5 0.0 0.5 1.0 1.5 2.5 2.0 1.5 1.0 0.5 0.0 0.5 1.0 3 2 1 0 1 2.5 2.0 1.5 1.0 0.5 0.… view at source ↗

**Figure 3.** Latent representation distributions of five example nodes from the Swiss roll graph using SIG-VAE (blue) and VGAE (red). SIG-VAE clearly infers more complex distributions that can be multi-modal, skewed, and with sharp and steep changes. This helps SIG-VAE to better represent the nodes in the latent space. variational parameters of SIG-VAE, we can derive a lower bound for the ELBO as follows (see the supplementary m… view at source ↗

Figures from the paper (1 more)

**Figure 4.** Figure 4: The nodes with multi-modal posteriors (red nodes) reside between different communities in Swiss Roll graph. We test the performances of SIG-VAE on different graph analytic tasks: 1) interpretability of SIG-VAE compared to VGAE, 2) link prediction in various real-world…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

42 extracted references · 26 canonical work pages

[1]

Martin Abadi, Ashish Agarwal, Paul Barham, Eugene Brevdo, Zhifeng Chen, Craig Citro, Greg S. Corrado, Andy Davis, Jeffrey Dean, Matthieu Devin, Sanjay Ghemawat, Ian Goodfellow, Andrew Harp, Geoffrey Irving, Michael Isard, Yangqing Jia, Rafal Jozefowicz, Lukasz Kaiser, Manjunath Kudlur, Josh Levenberg, Dan Mane, Rajat Monga, Sherry Moore, Derek Murray, Chr...

work page 2015
[2]

Distributed large-scale natural graph factorization

Amr Ahmed, Nino Shervashidze, Shravan Narayanamurthy, Vanja Josifovski, and Alexan- der J Smola. Distributed large-scale natural graph factorization. In Proceedings of the 22nd international conference on World Wide Web, pages 37–48. ACM, 2013. 9

work page 2013
[3]

Mixed membership stochastic blockmodels

Edoardo M Airoldi, David M Blei, Stephen E Fienberg, and Eric P Xing. Mixed membership stochastic blockmodels. Journal of Machine Learning Research, 9(Sep):1981–2014, 2008

work page 1981
[4]

Robust negative sam- pling for network embedding

Mohammadreza Armandpour, Patrick Ding, Jianhua Huang, and Xia Hu. Robust negative sam- pling for network embedding. In Proceedings of the AAAI Conference on Artiﬁcial Intelligence, volume 33, pages 3191–3198. AAAI, 2019

2019
[5]

Laplacian eigenmaps and spectral techniques for embedding and clustering

Mikhail Belkin and Partha Niyogi. Laplacian eigenmaps and spectral techniques for embedding and clustering. In Advances in neural information processing systems, pages 585–591, 2002

2002
[6]

Manifold regularization: A geometric framework for learning from labeled and unlabeled examples

Mikhail Belkin, Partha Niyogi, and Vikas Sindhwani. Manifold regularization: A geometric framework for learning from labeled and unlabeled examples. Journal of machine learning research, 7(Nov):2399–2434, 2006

work page 2006
[7]

Variational relevance vector machines

Christopher M Bishop and Michael E Tipping. Variational relevance vector machines. In Proceedings of the Sixteenth conference on Uncertainty in artiﬁcial intelligence, pages 46–53. Morgan Kaufmann Publishers Inc., 2000

work page 2000
[8]

Variational inference: A review for statisticians

David M Blei, Alp Kucukelbir, and Jon D McAuliffe. Variational inference: A review for statisticians. Journal of the American Statistical Association, 112(518):859–877, 2017

2017

Show all 42 references

[9]

Deep gaussian embedding of graphs: Unsuper- vised inductive learning via ranking

Aleksandar Bojchevski and Stephan Gunnemann. Deep gaussian embedding of graphs: Unsuper- vised inductive learning via ranking. In International Conference on Learning Representations, 2018

2018
[10]

Harp: Hierarchical representation learning for networks

Haochen Chen, Bryan Perozzi, Yifan Hu, and Steven Skiena. Harp: Hierarchical representation learning for networks. In Thirty-Second AAAI Conference on Artiﬁcial Intelligence, 2018

2018
[11]

Hyper- spherical variational auto-encoders

Tim R Davidson, Luca Falorsi, Nicola De Cao, Thomas Kipf, and Jakub M Tomczak. Hyper- spherical variational auto-encoders. arXiv preprint arXiv:1804.00891, 2018

2018 arXiv
[12]

Pygsp: Graph signal processing in python

Michael Defferrard, Lionel Martin, Rodrigo Pena, and Nathanael Perraudin. Pygsp: Graph signal processing in python. URL https://github.com/epfl-lts2/pygsp/
[13]

Convolutional neural networks on graphs with fast localized spectral ﬁltering

Michael Defferrard, Xavier Bresson, and Pierre Vandergheynst. Convolutional neural networks on graphs with fast localized spectral ﬁltering. In Advances in Neural Information Processing Systems, pages 3844–3852, 2016

2016
[14]

node2vec: Scalable feature learning for networks

Aditya Grover and Jure Leskovec. node2vec: Scalable feature learning for networks. In Proceedings of the 22nd ACM SIGKDD international conference on Knowledge discovery and data mining, pages 855–864. ACM, 2016

2016
[15]

Inductive representation learning on large graphs

Will Hamilton, Zhitao Ying, and Jure Leskovec. Inductive representation learning on large graphs. In Advances in Neural Information Processing Systems, pages 1024–1034, 2017

2017
[16]

An in- troduction to variational methods for graphical models

Michael I Jordan, Zoubin Ghahramani, Tommi S Jaakkola, and Lawrence K Saul. An in- troduction to variational methods for graphical models. Machine learning, 37(2):183–233, 1999

1999
[17]

Improved variational inference with inverse autoregressive ﬂow

Durk P Kingma, Tim Salimans, Rafal Jozefowicz, Xi Chen, Ilya Sutskever, and Max Welling. Improved variational inference with inverse autoregressive ﬂow. In Advances in neural informa- tion processing systems, pages 4743–4751, 2016

2016
[18]

Variational graph auto-encoders

Thomas N Kipf and Max Welling. Variational graph auto-encoders. arXiv preprint arXiv:1611.07308, 2016

2016 arXiv
[19]

Semi-supervised classiﬁcation with graph convolutional networks

Thomas N Kipf and Max Welling. Semi-supervised classiﬁcation with graph convolutional networks. In International Conference on Learning Representations, 2017

2017
[20]

Link-based classiﬁcation

Qing Lu and Lise Getoor. Link-based classiﬁcation. In Proceedings of the 20th International Conference on Machine Learning (ICML-03), pages 496–503, 2003

2003
[21]

Auxiliary deep generative models

Lars Maaloe, Casper Kaae Sonderby, Soren Kaae Sonderby, and Ole Winther. Auxiliary deep generative models. In International Conference on Machine Learning, pages 1445–1453, 2016. 10

2016
[22]

Finding community structure in networks using the eigenvectors of matrices

Mark EJ Newman. Finding community structure in networks using the eigenvectors of matrices. Physical review E, 74(3):036104, 2006

2006
[23]

Masked autoregressive ﬂow for density estimation

George Papamakarios, Theo Pavlakou, and Iain Murray. Masked autoregressive ﬂow for density estimation. In Advances in Neural Information Processing Systems, pages 2338–2347, 2017

2017
[24]

Fluid communities: a competitive, scalable and diverse community detection algorithm

Ferran Parés, Dario Garcia Gasulla, Armand Vilalta, Jonatan Moreno, Eduard Ayguadé, Jesús Labarta, Ulises Cortés, and Toyotaro Suzumura. Fluid communities: a competitive, scalable and diverse community detection algorithm. In International Conference on Complex Networks and th...

2017
[25]

Deepwalk: Online learning of social repre- sentations

Bryan Perozzi, Rami Al-Rfou, and Steven Skiena. Deepwalk: Online learning of social repre- sentations. In Proceedings of the 20th ACM SIGKDD international conference on Knowledge discovery and data mining, pages 701–710. ACM, 2014
[26]

Hierarchical variational models

Rajesh Ranganath, Dustin Tran, and David Blei. Hierarchical variational models. In Interna- tional Conference on Machine Learning, pages 324–333, 2016

2016
[27]

Variational inference with normalizing ﬂows

Danilo Jimenez Rezende and Shakir Mohamed. Variational inference with normalizing ﬂows. arXiv preprint arXiv:1505.05770, 2015

2015 arXiv
[28]

Collective classiﬁcation in network data

Prithviraj Sen, Galileo Namata, Mustafa Bilgic, Lise Getoor, Brian Galligher, and Tina Eliassi- Rad. Collective classiﬁcation in network data. AI magazine, 29(3):93, 2008

2008
[29]

Measuring isp topologies with rocketfuel

Neil Spring, Ratul Mahajan, and David Wetherall. Measuring isp topologies with rocketfuel. ACM SIGCOMM Computer Communication Review, 32(4):133–145, 2002

2002
[30]

Line: Large- scale information network embedding

Jian Tang, Meng Qu, Mingzhe Wang, Ming Zhang, Jun Yan, and Qiaozhu Mei. Line: Large- scale information network embedding. In Proceedings of the 24th International Conference on World Wide Web, pages 1067–1077. International World Wide Web Conferences Steering Committee, 2015

2015
[31]

Leveraging social media networks for classiﬁcation

Lei Tang and Huan Liu. Leveraging social media networks for classiﬁcation. Data Mining and Knowledge Discovery, 23(3):447–478, 2011

2011
[32]

Comparative assessment of large-scale data sets of protein–protein interactions

Christian V on Mering, Roland Krause, Berend Snel, Michael Cornell, Stephen G Oliver, Stanley Fields, and Peer Bork. Comparative assessment of large-scale data sets of protein–protein interactions. Nature, 417(6887):399, 2002

2002
[33]

Graphical models, exponential families, and variational inference

Martin J Wainwright, Michael I Jordan, et al. Graphical models, exponential families, and variational inference. Foundations and Trends in Machine Learning, 1(1–2):1–305, 2008

2008
[34]

Collective dynamics of small-world networks

Duncan J Watts and Steven H Strogatz. Collective dynamics of small-world networks. nature, 393(6684):440, 1998

1998
[35]

Deep learning via semi- supervised embedding

Jason Weston, Frederic Ratle, Hossein Mobahi, and Ronan Collobert. Deep learning via semi- supervised embedding. In Neural Networks: Tricks of the Trade , pages 639–655. Springer, 2012

2012
[36]

Representation learning on graphs with jumping knowledge networks

Keyulu Xu, Chengtao Li, Yonglong Tian, Tomohiro Sonobe, Ken-ichi Kawarabayashi, and Stefanie Jegelka. Representation learning on graphs with jumping knowledge networks. In International Conference on Machine Learning, pages 5453–5462, 2018

2018
[37]

How powerful are graph neural networks? In International Conference on Learning Representations, 2019

Keyulu Xu, Weihua Hu, Jure Leskovec, and Stefanie Jegelka. How powerful are graph neural networks? In International Conference on Learning Representations, 2019

2019
[38]

Revisiting semi-supervised learning with graph embeddings

Zhilin Yang, William W Cohen, and Ruslan Salakhutdinov. Revisiting semi-supervised learning with graph embeddings. arXiv preprint arXiv:1603.08861, 2016

2016 arXiv
[39]

Semi-implicit variational inference

Mingzhang Yin and Mingyuan Zhou. Semi-implicit variational inference. In International Conference on Machine Learning, pages 5660–5669, 2018

2018
[40]

Link prediction based on graph neural networks

Muhan Zhang and Yixin Chen. Link prediction based on graph neural networks. arXiv preprint arXiv:1802.09691, 2018. 11

2018 arXiv
[41]

Inﬁnite edge partition models for overlapping community detection and link prediction

Mingyuan Zhou. Inﬁnite edge partition models for overlapping community detection and link prediction. In Artiﬁcial Intelligence and Statistics, pages 1135–1143, 2015

2015
[42]

Semi-supervised learning using gaussian ﬁelds and harmonic functions

Xiaojin Zhu, Zoubin Ghahramani, and John D Lafferty. Semi-supervised learning using gaussian ﬁelds and harmonic functions. In Proceedings of the 20th International conference on Machine learning (ICML-03), pages 912–919, 2003. 12

2003

Pith tools

Reviewed August 14, 2026 · model on record in the stance chip above.

[1] [1]

Martin Abadi, Ashish Agarwal, Paul Barham, Eugene Brevdo, Zhifeng Chen, Craig Citro, Greg S. Corrado, Andy Davis, Jeffrey Dean, Matthieu Devin, Sanjay Ghemawat, Ian Goodfellow, Andrew Harp, Geoffrey Irving, Michael Isard, Yangqing Jia, Rafal Jozefowicz, Lukasz Kaiser, Manjunath Kudlur, Josh Levenberg, Dan Mane, Rajat Monga, Sherry Moore, Derek Murray, Chr...

work page 2015

[2] [2]

Distributed large-scale natural graph factorization

Amr Ahmed, Nino Shervashidze, Shravan Narayanamurthy, Vanja Josifovski, and Alexan- der J Smola. Distributed large-scale natural graph factorization. In Proceedings of the 22nd international conference on World Wide Web, pages 37–48. ACM, 2013. 9

work page 2013

[3] [3]

Mixed membership stochastic blockmodels

Edoardo M Airoldi, David M Blei, Stephen E Fienberg, and Eric P Xing. Mixed membership stochastic blockmodels. Journal of Machine Learning Research, 9(Sep):1981–2014, 2008

work page 1981

[4] [4]

Robust negative sam- pling for network embedding

Mohammadreza Armandpour, Patrick Ding, Jianhua Huang, and Xia Hu. Robust negative sam- pling for network embedding. In Proceedings of the AAAI Conference on Artiﬁcial Intelligence, volume 33, pages 3191–3198. AAAI, 2019

2019

[5] [5]

Laplacian eigenmaps and spectral techniques for embedding and clustering

Mikhail Belkin and Partha Niyogi. Laplacian eigenmaps and spectral techniques for embedding and clustering. In Advances in neural information processing systems, pages 585–591, 2002

2002

[6] [6]

Manifold regularization: A geometric framework for learning from labeled and unlabeled examples

Mikhail Belkin, Partha Niyogi, and Vikas Sindhwani. Manifold regularization: A geometric framework for learning from labeled and unlabeled examples. Journal of machine learning research, 7(Nov):2399–2434, 2006

work page 2006

[7] [7]

Variational relevance vector machines

Christopher M Bishop and Michael E Tipping. Variational relevance vector machines. In Proceedings of the Sixteenth conference on Uncertainty in artiﬁcial intelligence, pages 46–53. Morgan Kaufmann Publishers Inc., 2000

work page 2000

[8] [8]

Variational inference: A review for statisticians

David M Blei, Alp Kucukelbir, and Jon D McAuliffe. Variational inference: A review for statisticians. Journal of the American Statistical Association, 112(518):859–877, 2017

2017