mlsanityData to 7 Oct 2026

Flow Network based Generative Models for Non-Iterative Diverse Candidate Generation

arXiv 2106.04399 ↗First posted 8 Jun 2021cs.LGcited by 384

Not on the map. Why?

Cited by 36 in Diffusion models, but placed by its own abstract, which didn’t fit.

Abstract (arXiv)

This paper is about the problem of learning a stochastic policy for generating an object (like a molecular graph) from a sequence of actions, such that the probability of generating an object is proportional to a given positive reward for that object. Whereas standard return maximization tends to converge to a single return-maximizing sequence, there are cases where we would like to sample a diverse set of high-return solutions. These arise, for example, in black-box function optimization when few rounds are possible, each with large batches of queries, where the batches should be diverse, e.g., in the design of new molecules. One can also see this as a problem of approximately converting an energy function to a generative distribution. While MCMC methods can achieve that, they are expensive and generally only perform local exploration. Instead, training a generative policy amortizes the cost of search during training and yields to fast generation. Using insights from Temporal Difference learning, we propose GFlowNet, based on a view of the generative process as a flow network, making it possible to handle the tricky case where different trajectories can yield the same final state, e.g., there are many ways to sequentially add atoms to generate some molecular graph. We cast the set of trajectories as a flow and convert the flow consistency equations into a learning objective, akin to the casting of the Bellman equations into Temporal Difference methods. We prove that any global minimum of the proposed objectives yields a policy which samples from the desired distribution, and demonstrate the improved performance and diversity of GFlowNet on a simple domain where there are many modes to the reward function, and on a molecule synthesis task.

Key facts. Cited by 384 papers in the map (data through 7 Oct 2026). Builds on 26 of its 50 references that are in the map. First posted on arXiv 8 Jun 2021.

Builds on

26 of 50 references.

References · 26

  1. 2019PyTorch: An Imperative Style, High-Performance Deep Learning Librarycited by 20,160
  2. 2017Proximal Policy Optimization Algorithmscited by 14,441
  3. 2018Soft Actor-Critic: Off-Policy Maximum Entropy Deep Reinforcement Learning with a Stochastic Actorcited by 4,635
  4. 2017Neural Message Passing for Quantum Chemistrycited by 3,460
  5. 2015Variational Inference with Normalizing Flowscited by 2,161
  6. 2017Reinforcement Learning with Deep Energy-Based Policiescited by 659
  7. 2018Junction Tree Variational Autoencoder for Molecular Graph Generationcited by 647
  8. 2020Masked Label Prediction: Unified Message Passing Model for Semi-Supervised Classificationcited by 500
  9. 2019BatchBALD: Efficient and Diverse Batch Acquisition for Deep Bayesian Active Learningcited by 428
  10. 2018MolGAN: An implicit generative model for small molecular graphscited by 392
  11. 2020GraphAF: a Flow-based Autoregressive Model for Molecular Graph Generationcited by 301
  12. 2021GraphDF: A Discrete Flow Model for Molecular Graph Generationcited by 177
14 more
  1. 2020Implicit Under-Parameterization Inhibits Data-Efficient Deep Reinforcement Learningcited by 134
  2. 2021MARS: Markov Molecular Sampling for Multi-objective Drug Discoverycited by 125
  3. 2021Oops I Took A Gradient: Scalable Sampling for Discrete Distributionscited by 92
  4. 2019MolecularRNN: Generating realistic molecular graphs with optimized propertiescited by 80
  5. 2020Learning To Navigate The Synthetically Accessible Chemical Space Using Reinforcement Learningcited by 57
  6. 2020Interference and Generalization in Temporal Difference Learningcited by 41
  7. 2020Learning Discrete Energy-based Models via Auxiliary-variable Local Explorationcited by 29
  8. 2019Autoregressive Energy Machinescited by 23
  9. 2019Discrete Object Generation with Reversible Inductive Constructioncited by 16
  10. 2015Data Generation as Sequential Decision Makingcited by 15
  11. 2017Calibrating Energy-based Generative Adversarial Networkscited by 13
  12. 2017Generating Focussed Molecule Libraries for Drug Discovery with Recurrent Neural Networkscited by 12
  13. 2020Learning to Solve Network Flow Problems via Neural Decodingcited by 9
  14. 2019Approximate Inference in Discrete Distributions with Monte Carlo Tree Search and Value Functionscited by 6

Built on by

Cited by year (2026 to 7 Oct)

Citing papers per year: 2021: 4; 2022: 37; 2023: 73; 2024: 88; 2025: 97; 2026: 85.

384 papers.

Sort:

By area:

  1. 2022Diffusion Models: A Comprehensive Survey of Methods and Applicationscited by 1,107
  2. 2021GFlowNet Foundationscited by 286
  3. 2022Trajectory balance: Improved credit assignment in GFlowNetscited by 244
  4. 2022Graph neural networks for materials science and chemistrycited by 181
  5. 2022Biological Sequence Design with GFlowNetscited by 172
  6. 2023Protein Design with Guided Discrete Diffusioncited by 154
  7. 2022Bayesian Structure Learning with Generative Flow Networkscited by 153
  8. 2022Learning GFlowNets from partial episodes for improved convergence and stabilitycited by 129
  9. 2022Generative Flow Networks for Discrete Probabilistic Modelingcited by 113
  10. 2025Self-Evolving Curriculum for LLM Reasoningcited by 107
  11. 2023Policy Representation via Diffusion Probability Model for Reinforcement Learningcited by 105
  12. 2022GFlowNets and variational inferencecited by 103
38 more
  1. 2024Self-Improvement in Language Models: The Sharpening Mechanismcited by 102
  2. 2022Multi-Objective GFlowNetscited by 100
  3. 2023Consciousness in Artificial Intelligence: Insights from the Science of Consciousnesscited by 99
  4. 2023A theory of continuous generative flow networkscited by 98
  5. 2022Exploring Chemical Space with Score-based Out-of-distribution Generationcited by 95
  6. 2023Amortizing intractable inference in large language modelscited by 94
  7. 2023Better Training of GFlowNets with Local Credit and Incomplete Trajectoriescited by 83
  8. 2024Curiosity-driven Red-teaming for Large Language Modelscited by 83
  9. 2023Towards Understanding and Improving GFlowNet Trainingcited by 74
  10. 2024Preserving Diversity in Supervised Fine-Tuning of Large Language Modelscited by 68
  11. 2023Robust Scheduling with GFlowNetscited by 67
  12. 2023A Hitchhiker's Guide to Geometric GNNs for 3D Atomic Systemscited by 65
  13. 2023GFlowNets for AI-Driven Scientific Discoverycited by 63
  14. 2022Less Data, More Knowledge: Building Next Generation Semantic Communication Networkscited by 61
  15. 2023Generative Flow Networks as Entropy-Regularized RLcited by 60
  16. 2024A Review of Modern Recommender Systems Using Generative Models (Gen-RecSys)cited by 58
  17. 2024A Diffusion Model Framework for Unsupervised Neural Combinatorial Optimizationcited by 58
  18. 2023Diffusion Generative Flow Samplers: Improving learning signals through partial trajectory optimizationcited by 57
  19. 2024Learning diverse attacks on large language models for robust red-teaming and safety tuningcited by 54
  20. 2024Ant Colony Sampling with GFlowNets for Combinatorial Optimizationcited by 53
  21. 2022A-NeSI: A Scalable Approximate Method for Probabilistic Neurosymbolic Inferencecited by 50
  22. 2023Joint Bayesian Inference of Graphical Structure and Parameters with a Single Generative Flow Networkcited by 50
  23. 2022Generative Augmented Flow Networkscited by 48
  24. 2023Sample-efficient Multi-objective Molecular Optimization with GFlowNetscited by 48
  25. 2024Improved off-policy training of diffusion samplerscited by 47
  26. 2023Local Search GFlowNetscited by 45
  27. 2023GFlowNet-EM for learning compositional latent variable modelscited by 40
  28. 2023Let the Flows Tell: Solving Graph Combinatorial Optimization Problems with GFlowNetscited by 40
  29. 2024Discrete Probabilistic Inference as Control in Multi-path Environmentscited by 40
  30. 2022A Variational Perspective on Generative Flow Networkscited by 39
  31. 2024RGFN: Synthesizable Molecular Generation Using GFlowNetscited by 39
  32. 2025FlowRL: Matching Reward Distributions for LLM Reasoningcited by 38
  33. 2024Genetic-guided GFlowNets for Sample Efficient Molecular Optimizationcited by 36
  34. 2021From Machine Learning to Robotics: Challenges and Opportunities for Embodied Intelligencecited by 35
  35. 2022Group SELFIES: A Robust Fragment-Based Molecular String Representationcited by 32
  36. 2023Stochastic Generative Flow Networkscited by 32
  37. 2023Thompson sampling for improved exploration in GFlowNetscited by 31
  38. 2025Diffusion Tree Sampling: Scalable inference-time alignment of diffusion modelscited by 31

50 most cited of 384 All 384 (JSON)