Generative Flow Networks

Also known as GFlowNet

Generative Flow Networks (GFlowNets) build objects through sequences of actions and learn to sample completed objects in proportion to their rewards, offering an alternative to finding only the single best solution. Common training methods include flow matching, detailed balance, and trajectory balance, with replay, policy-gradient updates, and learned action abstractions used to improve exploration and efficiency. Applications include molecule and crystal design, graph generation, language-model reasoning, and strategic decision-making. Current work seeks to scale these methods while preserving diversity and stable training, and to establish when training objectives reliably guarantee that the learned sampler matches the target reward distribution.

Latest papers

57 papers10 papers in the last four weeks, up 233% on the four weeks before. 0.1% of all new papers.

Hierarchical Cooperative GFlowNets
Oct 25, 2026Pedro Dall'Antonia, Juan Belieni, Ana Carolina Erthal +3Generative Flow Networks Diversity-Aware Sampling
School of Applied Mathematics (EMaP), Getulio Vargas Foundation, Rio de Janeiro, Brazil · Department of Informatics Engineering, University of Coimbra, Coimbra, Portugal
SNACK: A Sequential Notation Framework for Probabilistic Graph Generation
Oct 14, 2026Hohyun Kim, Hyesung Kim, Min-hwan Oh +1Graph Generation Constrained Generative Modeling
Equal contribution · Seoul National University
Mixture of GFlowNets
Oct 12, 2026Omar G. Younis, Joseph D. Viviano, Chirayu Anant Haryan +3Generative Flow Networks Mode Collapse
Mila – Québec AI Institute · Intel · Université de Montréal
Towards Understanding the Training of GFlowNets: A Theoretical Perspective
Oct 9, 2026Rui Hu, Yu Chen, Zhuoran Li +1Generative Flow Networks Flow Matching
Institute for Interdisciplinary Information Sciences, Tsinghua University
Learning a Mixture of GFlowNets
Oct 6, 2026Tiago da Silva, Amauri H. Souza, Salem LahlouGenerative Flow Networks
MBZUAI, UAE · Avra, Brazil & Federal Institute of Cear´a, Brazil
Universality and Convergence of Generative Flows
Oct 4, 2026Leo BrunswicFlow Matching Generative Flow Networks
LLM Suite Team, JP Morgan Chase & Co.
Beyond Mode Collapse: Generating Diverse Synthetic Expert Conversations via Generative Flow Networks
Sep 29, 2026Sumit Asthana, Michael Ion, Kevyn Collins ThompsonGenerative Flow Networks Flow-Based Generative Modeling
University of Michigan Ann Arbor, MI, USA
FlowMap-OPD: Rollout--Kernel Separation for On-Policy Distillation of Few-Step Flow-Map Generators
Sep 29, 2026Zhiqi Li, Bo ZhuGenerative Flow Networks Flow Map Learning
Georgia Institute of Technology
One Proposal for Every Margin: Zero-Shot Amortized Sequential Importance Sampling for Binary Matrices
Sep 28, 2026Ruishuo Chen, Weijia Li, Xun Wang +3Importance Sampling Generative Flow Networks
Institute for Interdisciplinary Information Sciences, Tsinghua University · Department of Statistics and Data Science, Tsinghua University
CapField-OPD: Learning Continuous Capability Fields via Joint-Anchored Multi-Teacher On-Policy Distillation for Flow Models
Sep 28, 2026Pengyang Ling, Jiazi Bu, Yujie Zhou +6Generative Flow Networks Multi-Teacher Knowledge Distillation
University of Science and Technology of China · Shanghai Jiao Tong University · Fudan University +2
R $^2$ Flow: Recursive Self-Improvement via Recursive Skill Evolution
Sep 27, 2026Mingda Zhang, Qiang Huang, Yanjin Li +4LLM Agent Self-Improvement Generative Flow Networks
The Chinese University of Hong Kong, Shenzhen · Fudan University · University of Illinois at Urbana-Champaign +2
WTF?! Simulation-Free Reinforcement Learning with Wasserstein-Tilted Flow Maps
Sep 22, 2026Abbas Mammadov, Jerry Y. Huang, Justin Lin +5Reinforcement Learning Generative Flow Networks
University of Oxford · Carnegie Mellon University
Particle GFlowNets: Rethinking Generative Marginalization Models
Sep 10, 2026Tiago da Silva, Diego Mesquita, Salem LahlouMarkov Chain Monte Carlo Generative Flow Networks
MBZUAI · School of Applied Mathematics, Getulio Vargas Foundation
FlowNeg: GFlowNet-Guided Diverse Hard Negative Sampling for Knowledge Graph Embedding
Aug 24, 2026Ibne Farabi Shihab, Naoshin Anzum Hridi, Joyanta Jyoti MondalKG Embedding Hard Negative Mining
Department of Computer Science, Iowa State University, USA · Department of Computer Science and Engineering, BRAC University, Bangladesh · Department of Computer and Information Sciences, University of Delaware, USA
CrystalGRPO: Target-Aligned and Coverage-Preserving Reinforcement Learning for Flow-Based Crystal Structure Prediction
Aug 6, 2026Kaixiang Su, Hongfei Xue, Qiang ZhuCrystal Structure Prediction Group Relative Policy Optimization
Department of Computer Science, University of North Carolina at Charlotte · Department of Mechanical Engineering and Engineering Science, University of North Carolina at Charlotte · North Carolina Battery Complexity, Autonomous Vehicle and Electrification (BATT CAVE) Research Center
IFlowNets: Extending Generative Samplers to Learn Strategies in Incomplete Information Games
Aug 5, 2026Conor M. Artman, Nicholas Di, Scott PerkinsImperfect-Information Games Generative Flow Networks
AI Research Group, Lawrence Livermore National Laboratory · Deptartment of Statistics, Rice University · Strategic Competition Analysis Group, Lawrence Livermore National Laboratory
Information-Geometric Forward Policy Training in GFlowNets
Aug 4, 2026Yordan Raykov, Rodrigo VeigaInformation Geometry Natural Gradient Methods
School of Mathematical Sciences, University of Nottingham, Nottingham, UK
AlphaG-OPD: Reliability-Gated Sibling Counterfactuals for On-Policy Distillation in Symbolic Alpha Factor Discovery
Aug 2, 2026Yaoyu SuQuantitative Finance Generative Flow Networks
GFlowRL: Scaling Distribution-Matching RL to Large Language Models
Jul 15, 2026Xiaodong Liu, Michael Xu, Jack W. Stokes +3RL for Code Generation Reinforcement Learning
Microsoft Research
Self-conditioned Flow Map Language Models via Fixed-point Flows
Jul 1, 2026Jaehoon Yoo, Wonjung Kim, Floor Eijkelboom +4Normalizing Flows One-Step Generative Modeling
KAIST · University of Amsterdam · Carnegie Mellon University
Posterior Refinement: Fast Language Generation via Any-Order Flow Maps
Jun 23, 2026Manan Agarwal, Sheel Shah, Chanhyuk Lee +6Non-Autoregressive Text Generation Generative Flow Networks
Carnegie Mellon University · KAIST
FlowPipe: LLM-Enhanced Conditional Generative Flow Networks for Data Preparation Pipeline Construction
Jun 23, 2026Kunyu Ni, Lei Cao, Jie He +4Generative Flow Networks Flow-Based Generative Modeling
Ocean University of China China · University of Arizona USA · University of Science and Technology Beijing China +1
Latent Thought Flow: Efficient Latent Reasoning in Large Language Models
Jun 15, 2026Xiandong Zou, Jing Huang, Jianshu Li +1RL for Language Model Reasoning Efficient Language Model Reasoning
Singapore Management University · Ant Group
Proximal Policy Optimization for Amortized Discrete Sampling
Jun 14, 2026Anna Zykova-Myzina, Timofei Gritsaev, Daniil Tiapkin +1Molecular Graph Generation Generative Flow Networks
HSE University · Constructor University · CMAP, CNRS, École polytechnique, IPP
Flow-based Policy Adaptation without Policy Updates
Jun 4, 2026Luzhe Sun, Jingtian Ji, Haoran Chen +2Robot Policy Learning Robot Policy Adaptation
Toyota Technological Institute at Chicago · Stony Brook University
Your GFlowNet Secretly Learns an Optimal Transport Plan
Jun 4, 2026Ian Maksimov, Nikita Morozov, Denis Belomestny +1Generative Flow Networks Flow-Based Generative Modeling
1HSE University · 2Duisburg-Essen University
Scalable Inference-Time Annealing with Surrogate Likelihood Estimators
May 29, 2026Daniel Peñaherrera, Rishal Aggarwal, David Ryan KoesComputational Quantum Chemistry Generative Flow Networks
CMU-Pitt PhD Program in Computational Biology · Dept. of Computational & Systems Biology, University of Pittsburgh, Pittsburgh, PA 15260, USA
Conflict-Aware Additive Guidance for Flow Models under Compositional Rewards
May 20, 2026Xuehui Yu, Fucheng Cai, Meiyi Wang +2Constrained Generative Modeling Diffusion Model Guidance
Smart Systems Institute, National University of Singapore, Singapore · Faculty of Computing, Harbin Institute of Technology, Harbin, China · School of Computing, National University of Singapore, Singapore
Distribution Matching Distillation without Fake Score Network
May 19, 2026Youngjoong Kim, Deokyeong Lee, Jaesik ParkDiffusion Model Distillation Few-Step Diffusion Sampling
Department of Computer Science and Engineering, Seoul National University · Department of Computer Science and Engineering, Sogang University Seoul, Republic of Korea
CAB: Accelerating Flow and Diffusion Sampling via Rectification and Corrected Adams-Bashforth
May 16, 2026Anuska Roy, Pravin NairFew-Step Diffusion Sampling Generative Flow Networks
Department of Electrical Engineering Indian Institute of Technology Madras, Chennai, India

Oct 25, 2026J/K move · Enter open · S save

Hierarchical Cooperative GFlowNets

Pedro Dall'Antonia, Juan Belieni, Ana Carolina Erthal, Saullo H. G. de Oliveira+2

School of Applied Mathematics (EMaP), Getulio Vargas Foundation, Rio de Janeiro, Brazil · Department of Informatics Engineering, University of Coimbra, Coimbra, Portugal

Generative Flow Networks Diversity-Aware Sampling

Open paper AskPDF arXiv

Generative Flow Networks (GFlowNets) are trained to sample objects from compositional spaces with probability proportional to a reward function, but ineffective exploration can cause learned samplers to concentrate on a few high-reward regions. An intuitive remedy is to train multiple GFlowNets independently in parallel and combine their samples. However, this does not prevent multiple models from collapsing onto the same high-reward areas, leaving large regions of the state space unexplored. With this in mind, we propose Hierarchical Cooperative GFlowNets (HCG), an approach that dynamically partitions the search space to coordinate discovery without relying on hand-crafted boundaries. HCG imposes an ordering in which the highest-ranked model may explore the entire space, while lower-ranked models are progressively pushed toward regions not covered by higher-ranked ones. We provide a theoretical characterization of how specific learning objectives can shift probability mass away from designated areas, instantiating this mechanism within our architecture. A shared classifier coordinates the hierarchy by partitioning the search space, keeping the GFlowNets independently parameterized and enabling asynchronous training. Furthermore, we derive a normalizer-weighted aggregation rule for the ensemble and prove that it recovers the target reward-proportional distribution under the stated assumptions. Finally, experiments across multimodal distributions, sequence generation, combinatorial optimization, and biological sequence design show that HCG improves mode coverage and target approximation, translating increases in the number of coordinated samplers into broader, complementary exploration.