Latest Flow Matching Research Papers
The newest Flow Matching papers from across the field — arXiv, NeurIPS, CVPR, Nature, and more — refreshed daily and ranked by relevance. Distill AI tracks Flow Matching so you don’t have to: get the standout work delivered to your inbox every morning, with 2-sentence summaries and the option to chat with any paper.
Get the latest Flow Matching papers in your inbox — free →Recent papers
- Unsupervised Domain Adaptation For Enhanced Radiometer Image Precipitation Estimation Using Conditional Flow MatchingVictor Enescu, Assaad Zeghina, Matthieu Meignin, Nicolas Viltard et al. · HAL (Le Centre pour la Comm... · Sep 13, 2026
International audience...
- ROMS-IMLE: A Minimalist Approach to Competitive Single-Step Generative ModellingChirag Vashist, Ke Li · arXiv · Jul 21, 2026
Generative models have undergone many generations of evolution, from VAEs/GANs to diffusion/flow matching. Along the way, the underlying techniques have become more complicated and various beliefs about what drives strong empirical performa…
- A Shortcut to Statistically Steady-State Turbulence with Flow MatchingGianluca Galletti, Gerald Gutenbrunner, William Hornsby, Lorenzo Zanisi et al. · arXiv · Jul 14, 2026
Many nonlinear physical systems exhibit an initial transient phase in which perturbations grow before nonlinear interactions lead to a statistically steady state. While this saturated regime is of primary interest, direct numerical simulati…
- Guidance Breaks the Fitted Operator: A Terminal-Fitted Repair for Classifier-Free GuidanceShiheng Zhang · arXiv · Jul 8, 2026
Classifier-free guidance (CFG) is the standard way to strengthen class-conditioning in diffusion and flow-matching samplers, yet at large guidance it oversaturates and destabilizes, symptoms practitioners suppress with more steps or limited…
- Training-Free Acceleration for Vision-Language-Action Models with Action Caching and RefinementRyuji Oi, Hikari Otsuka, Kosuke Matsushima, Yuki Ichikawa et al. · arXiv · Jul 7, 2026
Vision-Language-Action (VLA) models have emerged as a promising approach for generalizable robotic manipulations. In particular, flow matching-based VLA models have shown remarkable success due to their capability to generate precise and sm…
- TriA Pipeline: A Large-Scale Automatic Audio Annotation Pipeline For Audio Classification In Specific ScenariosHong Lyu, Mingru Yang, Qianhua He, Yanxiong Li et al. · arXiv · Jul 7, 2026
There are some datasets of varying scales for audio classification (AC) applied to different tasks. However, annotated data is limited for most scenarios, such as domestic environments. To address this challenge, we propose an $\textbf{A}$u…
- The Fundamental Limits of Valid Transport Map EstimationSivaraman Balakrishnan · arXiv · Jun 29, 2026
Many modern generative modeling methods, including diffusion models, normalizing flows, and flow matching, estimate transport maps or plans between distributions without explicitly targeting an optimal transport (OT) map. In applications li…
- MedPCFM: Improving Medical Point Cloud Completion by Integrating Point Transformers and Flow MatchingKamil Kwarciak, Marek Wodzinski · arXiv · Jun 23, 2026
Medical point cloud completion is important for anatomical reconstruction and downstream clinical workflows, yet generative modeling in this setting remains insufficiently studied. We investigate completion through continuous-time generativ…
- Hedgementation = Hedgerow Segmentation: A Remote Sensing BenchmarkNathan Senyard, Salem Hamdani, Astrid Zhang, Derek Wang et al. · arXiv · Jun 22, 2026
We propose Hedgementation: a new benchmark to evaluate machine learning models for hedgerow mapping from remote sensing data at country scale and 10m$^2$ spatial resolution. We combine and harmonize multiple remote sensing data products and…
- Scaling Linear Mode Connectivity and Merging to Billion Parameter Pretrained TransformersTianyi Li, Zhiqiang Shen · arXiv · Jun 22, 2026
Linear mode connectivity (LMC) provides a promising foundation for understanding and merging independently trained neural networks, but existing methods typically optimize the interpolation path from only one model endpoint, limiting their …
- Claw-SWE-Bench: A Benchmark for Evaluating OpenClaw-style Agent Harnesses on Coding TasksMengyu Zheng, Kai Han, Boxun Li, Haiyang Xu et al. · arXiv · Jun 10, 2026
General-purpose agents such as OpenClaw are increasingly used as autonomous tool users, but their coding ability is difficult to measure under SWE-bench: a generic agent does not by itself satisfy the clean Docker workspace, patch, and pred…
- PianoKontext: Expressive Performance Rendering from Deadpan ContextDmitrii Gavrilev · arXiv · Jun 10, 2026
Expressive performance rendering (EPR) aims to generate realistic performances constrained on sequences of notes. However, flow matching audio editing models manipulate only synchronized music samples of the same duration, limiting their un…
- OncoTraj: a public benchmark for longitudinal resistance prediction in EGFR-mutant non-small-cell lung cancer on osimertinibAbhijoy Sarkar, Aarchi Singh Thakur · arXiv · Jun 9, 2026
Resistance to first-line osimertinib in EGFR-mutant non-small-cell lung cancer (NSCLC) is the canonical example of predictable clonal evolution under therapeutic pressure, yet no public benchmark exists for training or evaluating computatio…
- Learning Manifold Data with Flow MatchingSophia Pi, Mingcheng Lu, Jerry Yao-Chieh Hu, Maojiang Su et al. · SPIGM @ ICML Poster · May 30, 2026
We study flow-matching transformers when data lie on a low-dimensional manifold. Our key insight is a flow decomposition that splits motion along the manifold from motion off the manifold. The scheme works for first- and higher-order flow m…
- Learning Manifold Data with Flow MatchingSophia Pi, Mingcheng Lu, Jerry Yao-Chieh Hu, Maojiang Su et al. · ICML 2026 FoGen Workshop Poster · May 26, 2026
We study flow-matching transformers when data lie on a low-dimensional manifold. Our key insight is a flow decomposition that splits motion along the manifold from motion off the manifold. The scheme works for first- and higher-order flow m…
- Learning Manifold Data with Flow MatchingSophia Pi, Mingcheng Lu, Maojiang Su, Weimin Wu et al. · ICML 2026 regular · Apr 30, 2026
We study flow-matching transformers when data lie on a low-dimensional manifold. Our key insight is a flow decomposition that splits motion along the manifold from motion off the manifold. The scheme works for first and higher-order flow …
- Low-Pass Flow MatchingFrancesco M. Ruscio, T. Konstantin Rusch · ICLR 2026 DeLTa Workshop Poster · Mar 3, 2026
Flow Matching typically relies on white noise sources, a choice often misaligned with the power spectra of natural data, which tend to decay with frequency. To address this, we introduce $\textbf{Low-Pass Flow Matching}$, a variant of Flow …
- Flow Matching Policy GradientsDavid McAllister, Songwei Ge, Brent Yi, Chung Min Kim et al. · ICLR 2026 Poster · Jan 26, 2026
Flow-based generative models, including diffusion models, excel at modeling continuous distributions in high-dimensional spaces. In this work, we introduce Flow Policy Optimization (FPO), a simple on-policy reinforcement learning algorithm …
- Topological Flow MatchingKacper Wyrwal, Ismail Ilkan Ceylan, Alexander Tong · ICLR 2026 Poster · Jan 26, 2026
Flow matching is a powerful generative modeling framework, valued for its simplicity and strong empirical performance. However, its standard formulation treats signals on structured spaces---such as fMRI data on brain graphs---as points in …
- Active Flow MatchingYashvir Singh Grewal, Edwin V. Bonilla, Thang D Bui · AIML-CEB 2025 Oral · Nov 12, 2025
Discrete diffusion and flow matching excel at capturing epistatic structure in protein fitness landscapes through parallel, iterative refinement. However, their implicit nature—sampling via learned dynamics without tractable densities—preve…
- Federated Flow MatchingZifan Wang, Anqi Dong, Mahmoud Selim, Michael M. Zavlanos et al. · Submitted to ICLR 2026 · Sep 19, 2025
Data today is decentralized, generated and stored across devices and institutions where privacy, ownership, and regulation prevent centralization. This motivates the need to train generative models directly from distributed data locally wit…
- Coefficients-Preserving Sampling for Reinforcement Learning with Flow MatchingFeng Wang, Zihao Yu · arXiv.org · Sep 7, 2025
Reinforcement Learning (RL) has recently emerged as a powerful technique for improving image and video generation in Diffusion and Flow Matching models, specifically for enhancing output quality and alignment with prompts. A critical step f…
- Flow Matching Policy GradientsDavid McAllister, Songwei Ge, Brent Yi, Chung Min Kim et al. · arXiv.org · Jul 28, 2025
Flow-based generative models, including diffusion models, excel at modeling continuous distributions in high-dimensional spaces. In this work, we introduce Flow Policy Optimization (FPO), a simple on-policy reinforcement learning algorithm …
- On the Closed-Form of Flow Matching: Generalization Does Not Arise from Target StochasticityQuentin Bertrand, Anne Gagneux, Mathurin Massias, R'emi Emonet · arXiv.org · Jun 4, 2025
Modern deep generative models can now produce high-quality synthetic samples that are often indistinguishable from real training data. A growing body of research aims to understand why recent methods, such as diffusion and flow matching tec…
- Physics-Constrained Flow Matching: Sampling Generative Models with Hard ConstraintsUtkarsh Utkarsh, Pengfei Cai, Alan Edelman, Rafael Gómez-Bombarelli et al. · arXiv.org · Jun 4, 2025
Deep generative models have recently been applied to physical systems governed by partial differential equations (PDEs), offering scalable simulation and uncertainty-aware inference. However, enforcing physical constraints, such as conserva…
- ReinFlow: Fine-tuning Flow Matching Policy with Online Reinforcement LearningTonghe Zhang, Chao Yu, Sichang Su, Yu Wang · arXiv.org · May 28, 2025
We propose ReinFlow, a simple yet effective online reinforcement learning (RL) framework that fine-tunes a family of flow matching policies for continuous robotic control. Derived from rigorous RL theory, ReinFlow injects learnable noise in…
- ReinFlow: Fine-tuning Flow Matching Policy with Online Reinforcement LearningTonghe Zhang, Chao Yu, Sichang Su, Yu Wang · Online Reinforcement Learning, Flow Matching Policy, Fine-tuning, Robot Learning · May 11, 2025
We propose ReinFlow, a simple yet effective online reinforcement learning (RL) framework that fine-tunes a family of flow matching policies for continuous robotic control. Derived from rigorous RL theory, ReinFlow injects learnable noise in…
- Flow-GRPO: Training Flow Matching Models via Online RLJie Liu, Gongye Liu, Jiajun Liang, Yangguang Li et al. · arXiv.org · May 8, 2025
We propose Flow-GRPO, the first method to integrate online policy gradient reinforcement learning (RL) into flow matching models. Our approach uses two key strategies: (1) an ODE-to-SDE conversion that transforms a deterministic Ordinary Di…
- Variational Rectified Flow MatchingPengsheng Guo, Alex Schwing · ICML 2025 poster · May 1, 2025
We study Variational Rectified Flow Matching, a framework that enhances classic rectified flow matching by modeling multi-modal velocity vector-fields. At inference time, classic rectified flow matching 'moves' samples from a source distrib…
- CellFlow enables generative single-cell phenotype modeling with flow matchingDominik Klein, J. S. Fleck, D. Bobrovskiy, Lea Zimmermann et al. · bioRxiv · Apr 17, 2025