Pith. sign in

REVIEW 6 minor 1 cited by

Variational Inference for Latent Variable Models in High Dimensions

T0 review · 0 major / 6 minor · reviewed 2026-08-07 · deepseek-v4-flash

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

Pith's one-line read This paper establishes exact validity thresholds for mean-field variational inference in LDA and MMSB.

desk verdict A genuinely new and careful piece of theory that pins down when mean-field VI works for LDA and MMSB; worth serious refereeing. read the letter →

arxiv 2506.01893 v2 pith:6SZ6E6NR submitted 2025-06-02 math.ST math.PRstat.MLstat.TH

Chenyang Zhong , Sumit Mukherjee , Bodhisattva Sen This is my paper
classification math.STmath.PRstat.MLstat.TH MSC 62F1562C1060F10
keywords variationalinferencemean-fieldapproximationlatentDirichletallocationmixedmembershipstochasticblockmodelcollapsedposteriornonlinearlargedeviationspartiallygrouped
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 answer a practical question: when can a fast mean-field variational approximation be trusted to stand in for an intractable posterior? It develops a general framework that bounds the KL divergence between the variational posterior and the true posterior by a smoothness term plus a complexity term, then applies it to two benchmark models. For latent Dirichlet allocation, it proves that the per-word approximation error vanishes exactly when DK=o(n), where D is the number of documents, K the number of topics, and n the total number of words, and a matching lower bound shows the condition is necessary. For the mixed membership stochastic blockmodel, it proves that the usual fully factorized approximation can fail, and that a partially grouped version, which pairs the two membership draws attached to each directed edge, is accurate exactly when K=o(n). If these claims are right, practitioners get precise certificates for when variational inference is safe and guaranteed bounds on the ELBO as a proxy for log evidence.

What carries the argument

The object that carries the argument is the collapsed Hamiltonian f(Z), the log of the unnormalized posterior after global latent variables are integrated out; its partition function S_{n,K} is the normalizing constant of the collapsed posterior. Lemma 2.1 converts the MFVI gap into the identity D_KL(Q̂ ∥ P(Z|X)) = log S_{n,K} − sup_y {E_{Q_y}[f(Z)] − I(y)}, where Q_y ranges over product distributions on the local variables and I(y) is their KL divergence to the base measure. The proof then approximates this free energy by replacing f with a twice-differentiable extension F(y) on the continuous simplex; the error is split into two terms Δ1(y) and Δ2(y) coming from a two-term Taylor expansion. Their first and second moments are controlled through local Hessian bounds in row-wise and multiplicative neighborhoods, and a new probability lemma converts those moment bounds into a high-probability bound. A separate step uses the Gibbs variational principle to reduce full VI to collapsed VI, which is what lets the LDA and MMSB rates come out sharp.

What would settle it

In the symmetric LDA example of Theorem 1.2, set α=1/2, use a uniform topic-word matrix, take equal document lengths, and choose D and K so that DK = c n for a fixed c>0; if the minimized mean-field KL divergence per word decays to zero as n grows, the claimed exact regime is wrong, since the theorem predicts it stays bounded below by (c/5) log(1/c+2).

Watch

Extended reading notes

Core claim

The central claim is that, for Bayesian latent variable models with categorical local latent variables, mean-field variational inference approximates the true posterior well exactly when the total number of local categorical states is sublinear in the sample size, and the paper proves this by matching upper and lower bounds. In LDA, with n words, K topics, and vocabulary size V, the normalized KL gap is at most C(DK/n) log(n/DK+2) for every data set, so it vanishes exactly when DK=o(n); Theorem 1.2 shows that when DK is proportional to n, the gap stays bounded below by a constant multiple of the same expression. For the mixed membership stochastic blockmodel, the paper shows that the vanilla fully factorized mean-field posterior fails even in a simple symmetric example, and that partially grouped VI, which treats each directed pair of membership draws as one $K^{2}$-valued variable, achieves normalized KL gap at most C(K/n) log(n/K+2), with K=o(n) as the sharp regime. The general theorems supply the same kind of finite-sample KL and log-partition bounds for the whole class of models, not only for the two examples.

Load-bearing premise

The general theorems stand on the assumption that the collapsed Hamiltonian extends to a twice-differentiable function on the continuous simplex with bounded gradient and Hessian; without such a smooth extension, the error terms in the proof cannot be controlled and the upper bounds collapse.

Editorial extensions

If this is right

  • For LDA, practitioners can trust MFVI's posterior approximation and its ELBO as a proxy for log evidence only when the product DK is o(n); when DK is proportional to n, the per-word gap stays bounded below.
  • The LDA rates are optimal up to constants: the upper and lower bounds both scale as (DK/n) log(n/DK+2).
  • For MMSB, the standard fully factorized variational approximation should be replaced by partially grouped VI; treating each directed pair (Z_{i→j}, Z_{i←j}) as one categorical variable yields vanishing normalized KL when K=o(n).
  • The general framework gives non-asymptotic finite-sample bounds for any latent variable model in the stated class, including under model misspecification and for arbitrarily small hyperparameters, so the analysis covers both sparse and dense regimes.
  • Because the KL gap equals log p(X) minus the ELBO, the bounds also serve as certified error bounds for evidence approximation and for model-selection criteria based on the ELBO.

Reading between the lines

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

  • Beyond the paper, a natural conjecture is that other latent variable models obey a similar rule of thumb: validity of mean-field VI requires the product of the number of categorical states to be sublinear in sample size, so correlated topic models would plausibly need DK+KV=o(n), matching the paper's passing remark about a related Gaussian topic-model instability.
  • The MMSB result suggests a design heuristic: group local categorical variables that are strongly correlated under the posterior before applying mean field; the paper's simulations show Z_{i→j} and Z_{i←j} are highly correlated, and this grouping heuristic could be tested in other relational models.
  • The lower-bound examples use symmetric, exchangeable settings; with identifiable topics or communities, MFVI might remain accurate beyond the proved regime, so a useful stress test is whether the threshold shifts when topic-word matrices are sparse or the signal is strong.
  • Because the framework controls the log-partition function tightly, it could be repurposed for empirical Bayes estimation of hyperparameters, since the ELBO gap is now under explicit control; the paper lists this as a direction but does not develop it.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

0 major / 6 minor

Summary. This paper develops a general non-asymptotic framework for mean-field variational inference in Bayesian latent variable models with categorical local latent variables. Under explicit smoothness assumptions on the collapsed Hamiltonian F (Assumptions 2.1--2.3), Theorems 2.1 and 2.2 bound the log-partition function and the normalized KL divergence between the variational posterior and the true posterior. The framework is applied to two benchmark models. For LDA, Theorem 1.1 gives sup_X (1/n) D_KL(P̂ || P) ≤ C (DK/n) log(n/DK + 2), and Theorem 1.2 gives a matching lower bound in a symmetric example, establishing DK = o(n) as the exact regime of validity. For MMSB, Theorems 1.3 and 1.4 show that partially grouped VI achieves (1/n^2) D_KL ≤ C (K/n) log(n/K + 2) exactly in the regime K = o(n), while Theorem 1.5 exhibits a specific example where fully factorized VI has (1/n^2) D_KL ≥ δ. The proofs extend the nonlinear large deviations framework with local Hessian bounds, two-term Taylor expansions, and first- and second-moment control of the error terms.

Significance. If the results hold, they constitute a substantial contribution: this is the first finite-sample, exact-regime characterization of MFVI accuracy for LDA and MMSB. The upper bounds are uniform over data and hyperparameters, hold under model misspecification, and do not require posterior contraction or identifiability; the lower bounds match the upper rates up to constants. The general meta-theorems extend nonlinear large deviations with explicit constants and a new reduction from full VI to collapsed VI. The paper provides complete proofs in the appendices, including explicit verification of the key smoothness assumption, Assumption 2.3, for both LDA and MMSB. The main caveat is that the general framework's applicability beyond the two examples is conditional on verifying Assumption 2.3, which is a substantive but clearly stated premise.

minor comments (6)
  1. [Abstract and §1.1] The phrase 'arbitrary global latent variables' overstates the scope of the general framework: Assumptions 2.1--2.3 require the conditional posterior of θ given Z and X to factorize and require a twice continuously differentiable extension F(y) with explicit gradient and Hessian bounds. Please qualify this wording so readers know the general theorems are conditional on these substantive conditions.
  2. [Running headers, pages 2 and 7] 'Mixed mebership stochastic blockmodel' appears with a typo; it should be 'Mixed membership stochastic blockmodel'.
  3. [Section 3 heading] The section heading 'V ariational inference for latent Dirichlet allocation' contains an extra space; it should be 'Variational inference for latent Dirichlet allocation'.
  4. [Section 3, reduction to D = 1] The argument that the general D case follows from D = 1 via Jensen's inequality is compressed into a single sentence. A one-line derivation of sum_d n_d log(n_d/K + 2) ≤ n log(n/(DK) + 2) would improve transparency for readers who want to verify the scaling in D.
  5. [§1.3.2 and abstract] Theorem 1.5 demonstrates suboptimality of fully factorized VI in a specific two-community, fully observed example, but the abstract's wording could be read as a blanket statement that fully factorized VI is always suboptimal for MMSB. Please clarify that the claim is a counterexample showing a non-vanishing lower bound.
  6. [Reference [30]] The reference for Huang, Liu, and Chen appears to be missing the volume number; it should include 'Bayesian Analysis 15, 711--736.'

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the derivation is self-contained and the examples verify, rather than presuppose, the general assumptions.

full rationale

The paper's central derivation chain is not circular. Theorem 2.1 bounds the collapsed posterior's log-partition function and the KL divergence of MFVI in terms of the smoothness term E1 and complexity term E2, which are defined through explicit Hessian/covering quantities, not through the target KL divergence. Lemma 2.1 provides an identity relating the KL divergence to the log-partition function and a variational supremum, which is a standard decomposition rather than a fitted relation. The applications to LDA and MMSB are genuine verifications: Assumption 2.3 is checked explicitly via the log-gamma formulas (3.9) and (C.14), and the bounds on E1 and E2 are computed directly from derivatives of psi and the discretization sets, with no parameter fitted to the KL values being predicted. The lower bounds in Theorems 1.2 and 1.4 are constructed in symmetric examples to match the upper-bound rates; this is a tightness proof, not an input to the upper bounds. The paper cites prior work, including the authors' own [41], but that citation is contextual and not load-bearing for the main theorems. The condition DK=o(n) is derived from explicit finite-sample bounds, not imported from a self-citation. No step in the paper reduces a prediction to its inputs by definition or by statistical construction.

Assumptions & free parameters 0 free parameters · 9 assumptions · 0 invented entities

The central claim rests on no fitted parameters. All constants in the bounds are absolute. The model hyperparameters α, η, B are inputs to the prior or likelihood, not free parameters chosen to make the derivation work. The smoothness assumptions 2.1-2.3 are domain assumptions, verified for both applications. No new physical or probabilistic entities are introduced; partially grouped VI is an algorithmic grouping, not an invented entity.

assumptions (9)
  • domain assumption Assumption 2.1: the posterior density factorizes as exp(r0(Z) + Σ_j r_j(θ_j, Z))
    Required for the Gibbs variational principle step in Theorem 2.2; verified for LDA (Section 3) and MMSB (Appendix C).
  • domain assumption Assumption 2.2: there exists a smooth extension R(θ, y) of r(θ, z) that is twice continuously differentiable on the continuous simplex
    Needed to define the extended Hamiltonian and to differentiate F and R for Hessian bounds (Definitions 2.1-2.3).
  • domain assumption Assumption 2.3: F(y)=log ∫ exp(R(θ,y)) dν(θ) is twice continuously differentiable on the continuous simplex
    Central smoothness premise; if it fails, the Δ1 and Δ2 error bounds and Theorem 2.1 collapse.
  • domain assumption Hyperparameters α_l are uniformly bounded above by C0 in Theorems 1.1 and 1.3
    Used to control Stirling-approximation errors; allows arbitrarily small α and sparse regimes.
  • domain assumption For LDA, for each vocabulary r there is at least one topic ℓ with η_{ℓ,r} > 0
    Otherwise r can be removed from the vocabulary; ensures the base measure µ_i is well-defined.
  • domain assumption For MMSB, B is neither the zero matrix nor the matrix of all 1s
    Otherwise the observations are trivially all 0 or all 1, and the posterior is degenerate.
  • standard math Stirling approximation with uniform error (Lemma F.1)
    Used throughout to replace log Γ terms by x log x - x with controlled error.
  • standard math Gibbs variational principle: log ∫ e^{R} dν = sup_{ν'} [E_{ν'} R - D_KL(ν'||ν)]
    Used in the proof of Theorem 2.2 to relate the full Hamiltonian to F(y).
  • standard math Gronwall's inequality, Jensen's inequality, the data processing inequality, and convexity of log Γ on (0,∞)
    Used in the proofs of Propositions 2.1, 2.2, Theorem 2.1, and the LDA and MMSB applications.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Variational Inference for Latent Variable Models in High Dimensions." pith.science (2026). https://pith.science/paper/6SZ6E6NR

@misc{pith2026250601893,
  author       = {Pith},
  title        = {Pith review of: Variational Inference for Latent Variable Models in High Dimensions},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/6SZ6E6NR}},
  note         = {Machine review of arXiv:2506.01893}
}
read the original abstract

Variational inference (VI) is a popular method for approximating intractable posterior distributions in Bayesian inference and probabilistic machine learning. In this paper, we introduce a general framework for quantifying the statistical accuracy of mean-field variational inference (MFVI) for posterior approximation in Bayesian latent variable models with categorical local latent variables (and arbitrary global latent variables). Utilizing our general framework, we capture the exact regime where MFVI 'works' for the celebrated latent Dirichlet allocation model. Focusing on the mixed membership stochastic blockmodel, we show that the vanilla fully factorized MFVI, often used in the literature, is suboptimal. We propose a partially grouped VI algorithm for this model and show that it works, and derive its exact finite-sample performance. We further illustrate that our bounds are tight for both the above models. Our proof techniques, which extend the framework of nonlinear large deviations, open the door for the analysis of MFVI in other latent variable models.

Figures

Figures reproduced from arXiv: 2506.01893 by the authors.

Figure 1
Figure 1. Comparison of ELBO and scaled ELBO for fully factorized VI and partially grouped VI. [PITH_FULL_IMAGE:figures/full_fig_p032_1.png] view at source ↗

Discussion (0). Sign in to comment.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. Variational Bayes and Truncation approximations for Enriched Dirichlet process mixtures

    stat.ME 2026-03 conditional novelty 5.0 of 10

    VB-guided variable truncation of EDPMs has explicit L1 accuracy bounds and improves blocked-Gibbs mixing versus fixed or arbitrarily large truncations.

Reference graph

Works this paper leans on

62 extracted references · 55 canonical work pages · cited by 1 Pith paper

  1. [1]

    write newline

    " write newline "" before.all 'output.state := FUNCTION format.url url empty "" url if FUNCTION article output.bibitem format.authors "author" output.check author format.key output output.year.check new.block format.title "title" output.check new.block crossref missing format.jour.vol output format.article.crossref output.nonnull format.pages output if ne...

  2. [2]

    ( 2018 )

    barticle [author] Abbe , Emmanuel E. ( 2018 ). Community detection and stochastic block models: recent developments . Journal of Machine Learning Research 18 1--86 . barticle

    work page 2018

  3. [3]

    barticle [author] Airoldi , Edoardo M E. M. , Blei , David M D. M. , Fienberg , Stephen E S. E. Xing , Eric P E. P. ( 2008 ). Mixed membership stochastic blockmodels . Journal of Machine Learning Research 9 1981--2014 . barticle

    work page 2008

  4. [4]

    binproceedings [author] Airoldi , Edoardo M E. M. , Blei , David M D. M. , Fienberg , Stephen E S. E. , Xing , Eric P E. P. Jaakkola , Tommi T. ( 2006 ). Mixed membership stochastic block models for relational data with application to protein-protein interactions . In Proceedings of the International Biometrics Society Annual Meeting 15 1 . binproceedings

    work page 2006

  5. [5]

    , Ridgway , James J

    barticle [author] Alquier , Pierre P. , Ridgway , James J. Chopin , Nicolas N. ( 2016 ). On the properties of variational approximations of G ibbs posteriors . Journal of Machine Learning Research 17 1-41 . barticle

    work page 2016

  6. [6]

    Lacker , Daniel D

    barticle [author] Arnese , Manuel M. Lacker , Daniel D. ( 2024 ). Convergence of coordinate ascent variational inference for log-concave measures via optimal transport . arXiv preprint arXiv:2404.08792 . barticle

    arXiv 2024

  7. [7]

    ( 2021 )

    barticle [author] Augeri , Fanny F. ( 2021 ). A transportation approach to the mean-field approximation . Probability Theory and Related Fields 180 1--32 . barticle

    work page 2021

  8. [8]

    , Pati , Debdeep D

    barticle [author] Bhattacharya , Anirban A. , Pati , Debdeep D. Yang , Yun Y. ( 2025 ). On the convergence of coordinate ascent variational inference . The Annals of Statistics 53 929--962 . barticle

    work page 2025

Show all 62 references
  1. [9]

    , Choi , David D

    barticle [author] Bickel , Peter P. , Choi , David D. , Chang , Xiangyu X. Zhang , Hai H. ( 2013 ). Asymptotic normality of maximum likelihood and its variational approximation for stochastic blockmodels . The Annals of Statistics 41 1922--1943 . barticle

    2013

  2. [10]

    , Bunea , Florentina F

    barticle [author] Bing , Xin X. , Bunea , Florentina F. , Strimas-Mackey , Seth S. Wegkamp , Marten M. ( 2022 ). Likelihood estimation of sparse topic distributions in topic models and its applications to Wasserstein document distance calculations . The Annals of Statistics 50...

    2022

  3. [11]

    bbook [author] Bishop , Christopher M. C. M. ( 2006 ). Pattern recognition and machine learning . Information Science and Statistics . Springer, New York . 2247587 bbook

    2006

  4. [12]

    barticle [author] Blei , David M D. M. , Kucukelbir , Alp A. McAuliffe , Jon D J. D. ( 2017 ). Variational inference: A review for statisticians . Journal of the American Statistical Association 112 859--877 . barticle

    2017

  5. [13]

    binproceedings [author] Blei , David M D. M. Lafferty , John D J. D. ( 2005 ). Correlated topic models . In Proceedings of the 19th International Conference on Neural Information Processing Systems 147--154 . binproceedings

    2005

  6. [14]

    barticle [author] Blei , David M D. M. , Ng , Andrew Y A. Y. Jordan , Michael I M. I. ( 2003 ). Latent Dirichlet allocation . Journal of Machine Learning Research 3 993--1022 . barticle

    2003

  7. [15]

    , L'Huillier , Alice A

    barticle [author] Castillo , Isma \ A G l I. , L'Huillier , Alice A. , Ray , Kolyan K. Travis , Luke L. ( 2024 ). A variational Bayes approach to debiased inference for low-dimensional parameters in high-dimensional linear regression . arXiv preprint arXiv:2406.12659 . barticle

    2024 arXiv

  8. [16]

    Dembo , Amir A

    barticle [author] Chatterjee , Sourav S. Dembo , Amir A. ( 2016 ). Nonlinear large deviations . Advances in Mathematics 299 396--450 . barticle

    2016

  9. [17]

    , Handschy , Madeline M

    barticle [author] Chen , Wei-Kuo W.-K. , Handschy , Madeline M. Lerman , Gilad G. ( 2018 ). On the energy landscape of the mixed even p-spin model . Probability Theory and Related Fields 171 53--95 . barticle

    2018

  10. [18]

    bbook [author] Cover , Thomas M T. M. Thomas , Joy A J. A. ( 2012 ). Elements of Information Theory . John Wiley & Sons . bbook

    2012

  11. [19]

    barticle [author] Dunson , David B D. B. ( 2000 ). Bayesian latent variable models for clustered mixed outcomes . Journal of the Royal Statistical Society Series B: Statistical Methodology 62 355--366 . barticle

    2000

  12. [20]

    Rigon , Tommaso T

    barticle [author] Durante , Daniele D. Rigon , Tommaso T. ( 2019 ). Conditionally conjugate mean-field variational Bayes for logistic models . Statistical Science 34 472--485 . barticle

    2019

  13. [21]

    ( 2018 )

    barticle [author] Eldan , Ronen R. ( 2018 ). Gaussian-width gradient complexity, reverse log-Sobolev inequalities and nonlinear large deviations . Geometric and Functional Analysis 28 1548--1596 . barticle

    2018

  14. [22]

    , Stephens , Matthew M

    barticle [author] Falush , Daniel D. , Stephens , Matthew M. Pritchard , Jonathan K J. K. ( 2003 ). Inference of population structure using multilocus genotype data: linked loci and correlated allele frequencies . Genetics 164 1567--1587 . barticle

    2003

  15. [23]

    , Guan , Leying L

    barticle [author] Fan , Zhou Z. , Guan , Leying L. , Shen , Yandi Y. Wu , Yihong Y. ( 2023 ). Gradient flows for empirical Bayes in high-dimensional linear models . arXiv preprint arXiv:2312.12708 . barticle

    2023 arXiv

  16. [24]

    , Boyles , Levi L

    binproceedings [author] Foulds , James J. , Boyles , Levi L. , DuBois , Christopher C. , Smyth , Padhraic P. Welling , Max M. ( 2013 ). Stochastic collapsed variational Bayesian inference for latent Dirichlet allocation . In Proceedings of the 19th ACM SIGKDD international con...

    2013

  17. [25]

    , Ma , Zongming Z

    barticle [author] Gao , Chao C. , Ma , Zongming Z. , Zhang , Anderson Y A. Y. Zhou , Harrison H H. H. ( 2018 ). Community detection in degree-corrected block models . The Annals of Statistics 46 2153--2185 . barticle

    2018

  18. [26]

    Klopp , Olga O

    barticle [author] Gaucher , Solenne S. Klopp , Olga O. ( 2021 ). Optimality of variational inference for stochastic block model with missing links . Advances in Neural Information Processing Systems 34 19947--19959 . barticle

    2021

  19. [27]

    , Javadi , Hamid H

    binproceedings [author] Ghorbani , Behrooz B. , Javadi , Hamid H. Montanari , Andrea A. ( 2019 ). An instability in variational inference for topic models . In International Conference on Machine Learning 2221--2231 . PMLR . binproceedings

    2019

  20. [28]

    barticle [author] Gopalan , Prem K P. K. Blei , David M D. M. ( 2013 ). Efficient discovery of overlapping communities in massive networks . Proceedings of the National Academy of Sciences 110 14534--14539 . barticle

    2013

  21. [29]

    Dunson , David B D

    barticle [author] Gu , Yuqi Y. Dunson , David B D. B. ( 2023 ). Bayesian pyramids: I dentifiable multilayer discrete latent structure models for discrete data . Journal of the Royal Statistical Society Series B: Statistical Methodology 85 399--426 . barticle

    2023

  22. [30]

    ( 1982 )

    bbook [author] Hartman , Philip P. ( 1982 ). Ordinary differential equations , second ed. Birkh\"auser, Boston, MA . 658490 bbook

    1982

  23. [31]

    , Liu , Yan Y

    barticle [author] Huang , Weihong W. , Liu , Yan Y. Chen , Yuguo Y. ( 2020 ). Mixed Membership Stochastic Blockmodels for Heterogeneous Networks . Bayesian Analysis 15 711--736 . barticle

    2020

  24. [32]

    barticle [author] Jordan , Michael I M. I. , Ghahramani , Zoubin Z. , Jaakkola , Tommi S T. S. Saul , Lawrence K L. K. ( 1999 ). An introduction to variational methods for graphical models . Machine learning 37 183--233 . barticle

    1999

  25. [33]

    barticle [author] Karatsuba , Ekatherina A E. A. ( 2001 ). On the asymptotic representation of the Euler gamma function by Ramanujan . Journal of Computational and Applied Mathematics 135 225--240 . barticle

    2001

  26. [34]

    barticle [author] Ke , Zheng Tracy Z. T. Wang , Minzhe M. ( 2024 ). Using SVD for topic modeling . Journal of the American Statistical Association 119 434--449 . barticle

    2024

  27. [35]

    , Wang , Wei W

    barticle [author] Kim , Youngseok Y. , Wang , Wei W. , Carbonetto , Peter P. Stephens , Matthew M. ( 2024 ). A flexible empirical Bayes approach to multiple linear regression and connections with penalized regression . Journal of Machine Learning Research 25 1--59 . barticle

    2024

  28. [36]

    barticle [author] Laureate , Caitlin Doogan Poet C. D. P. , Buntine , Wray W. Linger , Henry H. ( 2023 ). A systematic review of the use of topic models for short text social media analysis . Artificial Intelligence Review 56 14223--14255 . barticle

    2023

  29. [37]

    Zanella , Giacomo G

    barticle [author] Lavenant , Hugo H. Zanella , Giacomo G. ( 2024 ). Convergence rate of random scan coordinate ascent variational inference under log-concavity . SIAM Journal on Optimization 34 3750--3761 . barticle

    2024

  30. [38]

    Rinaldo , Alessandro A

    barticle [author] Lei , Jing J. Rinaldo , Alessandro A. ( 2015 ). Consistency of spectral clustering in stochastic block models . The Annals of Statistics 215--237 . barticle

    2015

  31. [39]

    , Moss , Fabian C F

    barticle [author] Lieck , Robert R. , Moss , Fabian C F. C. Rohrmeier , Martin M. ( 2020 ). The Tonal Diffusion Model . Transactions of the International Society for Music Information Retrieval 3 153--164 . barticle

    2020

  32. [40]

    , Qiu , Jiaze J

    barticle [author] Mukherjee , Sumit S. , Qiu , Jiaze J. Sen , Subhabrata S. ( 2024 ). On Naive Mean-Field Approximation for high-dimensional canonical GLMs . arXiv preprint arXiv:2406.15247 . barticle

    2024 arXiv

  33. [41]

    , Sen , Bodhisattva B

    barticle [author] Mukherjee , Sumit S. , Sen , Bodhisattva B. Sen , Subhabrata S. ( 2023 ). A mean field approach to empirical Bayes estimation in high-dimensional linear regression . arXiv preprint arXiv:2309.16843 . barticle

    2023 arXiv

  34. [42]

    Sen , Subhabrata S

    barticle [author] Mukherjee , Sumit S. Sen , Subhabrata S. ( 2022 ). Variational inference in high-dimensional linear regression . Journal of Machine Learning Research 23 1--56 . barticle

    2022

  35. [43]

    barticle [author] Parker , Maria A M. A. , Valdez , Danny D. , Rao , Varun K V. K. , Eddens , Katherine S K. S. Agley , Jon J. ( 2023 ). Results and Methodological Implications of the Digital Epidemiology of Prescription Drug References Among Twitter Users: Latent Dirichlet Al...

    2023

  36. [44]

    , Bhattacharya , Anirban A

    binproceedings [author] Pati , Debdeep D. , Bhattacharya , Anirban A. Yang , Yun Y. ( 2018 ). On statistical optimality of variational Bayes . In International Conference on Artificial Intelligence and Statistics 1579--1588 . PMLR . binproceedings

    2018

  37. [45]

    barticle [author] Pritchard , Jonathan K J. K. , Stephens , Matthew M. Donnelly , Peter P. ( 2000 ). Inference of population structure using multilocus genotype data . Genetics 155 945--959 . barticle

    2000

  38. [46]

    Szab \'o , Botond B

    barticle [author] Ray , Kolyan K. Szab \'o , Botond B. ( 2022 ). Variational Bayes for high-dimensional linear regression with sparse priors . Journal of the American Statistical Association 117 1270--1281 . barticle

    2022

  39. [47]

    barticle [author] Snijders , Tom AB T. A. ( 2011 ). Statistical models for social networks . Annual Review of Sociology 37 131--153 . barticle

    2011

  40. [48]

    barticle [author] Snijders , Tom AB T. A. Nowicki , Krzysztof K. ( 1997 ). Estimation and prediction for stochastic blockmodels for graphs with latent block structure . Journal of Classification 14 75--100 . barticle

    1997

  41. [49]

    , Chu , Jing J

    barticle [author] Song , Fangda F. , Chu , Jing J. , Ma , Shuangge S. Wei , Yingying Y. ( 2024 ). Survival mixed membership blockmodel . Journal of the American Statistical Association 119 1647--1656 . barticle

    2024

  42. [50]

    binproceedings [author] Teh , Yee Whye Y. W. , Newman , David D. Welling , Max M. ( 2006 ). A collapsed variational Bayesian inference algorithm for latent Dirichlet allocation . In Proceedings of the 20th International Conference on Neural Information Processing Systems 1353-...

    2006

  43. [51]

    ( 2014 )

    barticle [author] Van Handel , Ramon R. ( 2014 ). Probability in high dimension . Lecture Notes (Princeton University) . barticle

    2014

  44. [52]

    Kumar , Sathish AP S

    barticle [author] Vayansky , Ike I. Kumar , Sathish AP S. A. ( 2020 ). A review of topic modeling methods . Information Systems 94 101582 . barticle

    2020

  45. [53]

    barticle [author] Wainwright , Martin J M. J. Jordan , Michael I M. I. ( 2008 ). Graphical models, exponential families, and variational inference . Foundations and Trends in Machine Learning 1 1--305 . barticle

    2008

  46. [54]

    Blei , David M D

    binproceedings [author] Wang , Chong C. Blei , David M D. M. ( 2011 ). Collaborative topic modeling for recommending scientific articles . In Proceedings of the 17th ACM SIGKDD international conference on Knowledge discovery and data mining 448--456 . binproceedings

    2011

  47. [55]

    , Yang , Wen W

    barticle [author] Wang , Xiaojuan X. , Yang , Wen W. , Yang , Yue Y. , He , Yizhou Y. , Zhang , Jun J. , Wang , Lusheng L. Hu , Lun L. ( 2022 ). PPISB : a novel network-based algorithm of predicting protein-protein interactions with mixed membership stochastic blockmodel . IEE...

    2022

  48. [56]

    Blei , David M D

    barticle [author] Wang , Yixin Y. Blei , David M D. M. ( 2019 ). Frequentist consistency of variational Bayes . Journal of the American Statistical Association 114 1147--1161 . barticle

    2019

  49. [57]

    barticle [author] Wang , Yuchung J Y. J. Wong , George Y G. Y. ( 1987 ). Stochastic blockmodels for directed graphs . Journal of the American Statistical Association 82 8--19 . barticle

    1987

  50. [58]

    barticle [author] Xing , Eric P E. P. , Fu , Wenjie W. Song , Le L. ( 2010 ). A state-space mixed membership blockmodel for dynamic network tomography . The Annals of Applied Statistics 4 535-566 . barticle

    2010

  51. [59]

    ( 2020 )

    barticle [author] Yan , Jun J. ( 2020 ). Nonlinear large deviations: Beyond the hypercube . The Annals of Applied Probability 30 812--846 . barticle

    2020

  52. [60]

    barticle [author] Zhang , Anderson Y A. Y. Zhou , Harrison H H. H. ( 2016 ). Minimax rates of community detection in stochastic block models . The Annals of Statistics 2252--2280 . barticle

    2016

  53. [61]

    barticle [author] Zhang , Anderson Y A. Y. Zhou , Harrison H H. H. ( 2020 ). Theoretical and computational guarantees of mean field variational inference for community detection . The Annals of Statistics 48 2575--2598 . barticle

    2020

  54. [62]

    Gao , Chao C

    barticle [author] Zhang , Fengshuo F. Gao , Chao C. ( 2020 ). Convergence rates of variational posterior distributions . The Annals of Statistics 48 2180--2207 . barticle

    2020

Pith tools

Reviewed August 7, 2026 · model on record in the stance chip above.