aiqbench analysis
This research proposes a novel framework for one-step generative modeling on finite state spaces by leveraging discrete Wasserstein geometry and reversible Markov kernels. The method amortizes complex transport dynamics into a latent-conditioned generator, enabling efficient inference while maintaining theoretical guarantees like KL dissipation during training. These findings establish a foundational approach for scaling generative models to structured discrete data without iterative sampling at inference time.
Discrete Wasserstein Flows for One-Step Generative Modeling
arXiv:2610.01355v1 Announce Type: cross Abstract: We introduce a new framework for one-step generative modelling on finite state spaces. To extend drifting beyond continuous domains, we use discrete Wasserstein geometry to define a target-relative KL gradient flow over the transitions of a reversible Markov kernel. We realize this probability flow at the particle level through Markov jumps and amortize the resulting transport updates into a latent-conditioned generator, so that the iterative dynamics are required only during training while inference remains one-step. In a controlled setting where the underlying distributions and transport dynamics can be computed exactly, we verify KL dissipation, consistency between the particle dynamics and the probability flow, and the predicted numerical scaling. We further show that a finite-capacity neural generator can track these exact transport targets while retaining one-step generation. These results validate the basic construction and provide a foundation for scaling Discrete Drifting to structured discrete data.
Abstract:We introduce a new framework for one-step generative modelling on finite state spaces. To extend drifting beyond continuous domains, we use discrete Wasserstein geometry to define a target-relative KL gradient flow over the transitions of a reversible Markov kernel. We realize this probability flow at the particle level through Markov jumps and amortize the resulting transport updates into a latent-conditioned generator, so that the iterative dynamics are required only during training while inference remains one-step. In a controlled setting where the underlying distributions and transport dynamics can be computed exactly, we verify KL dissipation, consistency between the particle dynamics and the probability flow, and the predicted numerical scaling. We further show that a finite-capacity neural generator can track these exact transport targets while retaining one-step generation. These results validate the basic construction and provide a foundation for scaling Discrete Drifting to structured discrete data.
Machine Learning (cs.LG); Artificial Intelligence (cs.AI)
arXiv:2610.01355 [cs.LG]
(or arXiv:2610.01355v1 [cs.LG] for this version)
https://doi.org/10.48550/arXiv.2610.01355 Focus to learn more arXiv-issued DOI via DataCite (pending registration)
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.
Full text under the publisher's open licence · 241 words
Related coverage
LC-GRPO: Bridging Train-Inference Gap for Flow-Based GRPO with Langevin Correction
arXiv · Aug 7, 2026
RW-Flow: One-Step Generation on Compact Manifolds via Wasserstein Gradient Flows
arXiv · Oct 1, 2026
Inference meta-monitoring for Amazon SageMaker AI endpoints with Amazon Quick
AWS · Jul 30, 2026
Latent-Kernel Discrete Flow Maps for Few-Step Generation
arXiv · Jul 31, 2026