mlsanityData to 7 Oct 2026

Embarrassingly Parallel GFlowNets

arXiv 2406.03288 ↗First posted 5 Jun 2024cs.LG stat.MLcited by 3

Not on the map. Why?

Abstract (arXiv)

GFlowNets are a promising alternative to MCMC sampling for discrete compositional random variables. Training GFlowNets requires repeated evaluations of the unnormalized target distribution or reward function. However, for large-scale posterior sampling, this may be prohibitive since it incurs traversing the data several times. Moreover, if the data are distributed across clients, employing standard GFlowNets leads to intensive client-server communication. To alleviate both these issues, we propose embarrassingly parallel GFlowNet (EP-GFlowNet). EP-GFlowNet is a provably correct divide-and-conquer method to sample from product distributions of the form R(·) propto R1(·) ... RN(·) -- e.g., in parallel or federated Bayes, where each Rn is a local posterior defined on a data partition. First, in parallel, we train a local GFlowNet targeting each Rn and send the resulting models to the server. Then, the server learns a global GFlowNet by enforcing our newly proposed aggregating balance condition, requiring a single communication step. Importantly, EP-GFlowNets can also be applied to multi-objective optimization and model reuse. Our experiments illustrate the EP-GFlowNets's effectiveness on many tasks, including parallel Bayesian phylogenetics, multi-objective multiset, sequence generation, and federated Bayesian structure learning.

Key facts. Cited by 3 papers in the map (data through 7 Oct 2026). Builds on 44 of its 62 references that are in the map. First posted on arXiv 5 Jun 2024.

Builds on

44 of 62 references.

References · 44

  1. 2016Communication-Efficient Learning of Deep Networks from Decentralized Datacited by 7,592
  2. 2018How Powerful are Graph Neural Networks?cited by 4,355
  3. 2019Fast Graph Representation Learning with PyTorch Geometriccited by 2,414
  4. 2018DAGs with NO TEARS: Continuous Optimization for Structure Learningcited by 745
  5. 2021Flow Network based Generative Models for Non-Iterative Diverse Candidate Generationcited by 384
  6. 2021GFlowNet Foundationscited by 286
  7. 2022Trajectory balance: Improved credit assignment in GFlowNetscited by 244
  8. 2023Reduce, Reuse, Recycle: Compositional Generation with Energy-Based Diffusion Models and MCMCcited by 216
  9. 2022Biological Sequence Design with GFlowNetscited by 172
  10. 2021Beware of the Simulated DAG! Causal Discovery Benchmarks May Be Easy To Gamecited by 159
  11. 2022Bayesian Structure Learning with Generative Flow Networkscited by 153
  12. 2021Parallel Bayesian Optimization of Multiple Noisy Objectives with Expected Hypervolume Improvementcited by 115
32 more
  1. 2022Generative Flow Networks for Discrete Probabilistic Modelingcited by 113
  2. 2021DiBS: Differentiable Bayesian Structure Learningcited by 105
  3. 2022GFlowNets and variational inferencecited by 103
  4. 2022Multi-Objective GFlowNetscited by 100
  5. 2023A theory of continuous generative flow networkscited by 98
  6. 2023Personalized Federated Learning with Feature Alignment and Classifier Collaborationcited by 97
  7. 2023Amortizing intractable inference in large language modelscited by 94
  8. 2022Deep End-to-end Causal Inferencecited by 93
  9. 2023Better Training of GFlowNets with Local Credit and Incomplete Trajectoriescited by 83
  10. 2023Towards Understanding and Improving GFlowNet Trainingcited by 74
  11. 2020VarGrad: A Low-Variance Gradient Estimator for Variational Inferencecited by 73
  12. 2023Robust Scheduling with GFlowNetscited by 67
  13. 2023Joint Bayesian Inference of Graphical Structure and Parameters with a Single Generative Flow Networkcited by 50
  14. 2022Generative Augmented Flow Networkscited by 48
  15. 2013Streaming Variational Bayescited by 46
  16. 2023GFlowNet-EM for learning compositional latent variable modelscited by 40
  17. 2023Let the Flows Tell: Solving Graph Combinatorial Optimization Problems with GFlowNetscited by 40
  18. 2020Synthetic Learning: Learn From Distributed Asynchronized Discriminator GAN Without Sharing Medical Image Datacited by 38
  19. 2023Stochastic Generative Flow Networkscited by 32
  20. 2021Towards Federated Bayesian Network Structure Learning with Continuous Optimizationcited by 31
  21. 2013Asymptotically Exact, Embarrassingly Parallel MCMCcited by 30
  22. 2023Distributional GFlowNets with Quantile Flowscited by 28
  23. 2023DynGFN: Towards Bayesian Inference of Gene Regulatory Networks with GFlowNetscited by 27
  24. 2021QLSD: Quantised Langevin stochastic dynamics for Bayesian federated learningcited by 26
  25. 2023Compositional Sculpting of Iterative Generative Processescited by 26
  26. 2020Federated Stochastic Gradient Langevin Dynamicscited by 16
  27. 2016Merging MCMC Subposteriors through Gaussian-Process Approximationscited by 13
  28. 2019Embarrassingly parallel MCMC using deep invertible transformationscited by 11
  29. 2022Parallel MCMC Without Embarrassing Failurescited by 11
  30. 2015Parallelizing MCMC with Random Partition Treescited by 10
  31. 2023Expert-Aided Causal Discovery of Ancestral Graphscited by 8
  32. 2020Learn distributed GAN with Temporary Discriminatorscited by 5

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2026: 3.

3 papers.

  1. 2026Signal from Structure: Exploiting Submodular Upper Bounds in Generative Flow Networkscited by 0
  2. 2026Stable-GFlowNet: Toward Diverse and Robust LLM Red-Teaming via Contrastive Trajectory Balancecited by 0
  3. 2026Path-dependent Discrete Amortized Inferencecited by 0