DrPO enables online preference optimization for deterministic one-step generators via non-parametric dipole updates from ranked samples plus base-model drift, without reward backpropagation.
hub
Sinkhorn-Drifting Generative Models.arXiv preprint arXiv:2603.12366, 2026a
13 Pith papers cite this work. Polarity classification is still indexing.
13
Pith papers citing it
hub tools
citation-role summary
background 3
citation-polarity summary
years
2026 13verdicts
UNVERDICTED 13roles
background 3polarities
background 3representative citing papers
A unified framework decomposes Wasserstein gradient flow velocity fields across f-divergences into a shared beta direction and divergence-specific weighting, enabling data-free one-step sampling.
SROT regularizes the OT transport plan toward a sliced OT reference, yielding better approximations of exact OT than entropic OT and improving on the sliced OT plan itself.
A new constrained gradient flow on the space of transport maps converges to the OT map and enables more stable and accurate training of convexity-constrained neural networks for learning Monge maps.
Derives continuous-time finite-particle convergence rates for a new conservative KDE-gradient drifting method and the non-conservative Laplace kernel method in one-step generative modeling.
W-Flow compresses a Wasserstein gradient flow defined via Sinkhorn divergence into a single-step neural generator, reporting 1.29 FID on ImageNet 256x256 with improved mode coverage.
DFP is a one-step generative policy using Wasserstein gradient flow on a drifting model backbone, with a top-K behavior cloning surrogate, that reaches SOTA on Robomimic and OGBench manipulation tasks.
SymDrift makes drifting models produce symmetry-invariant samples in one step via symmetrized coordinate drifts or G-invariant embeddings, outperforming prior one-shot baselines on molecular benchmarks and cutting compute by up to 40x.
The paper interprets GMD algorithms as limiting points of Wasserstein gradient flows on KL divergence with Parzen smoothing and on Sinkhorn divergence, while extending the approach to MMD, sliced Wasserstein, and GAN critics.
The paper proves identifiability of drifting fields for companion-elliptic kernels and shows that field convergence plus a C0 observable recovers weak convergence even without tightness.
RA-OT and OA-OT amortize optimal transport by regressing or optimizing sliced-OT Kantorovich potentials to approximate full OT plans efficiently across multiple measure pairs.
Drift Flow Matching connects direct transport maps from Drift Models with flow-based iterative refinement to enable adaptive computation in generative modeling.
A simplified one-step diffusion distillation uses pretrained teacher features directly for drifting loss plus a mode coverage term, achieving FID 1.58 on ImageNet-64 and 18.4 on SDXL.
citing papers explorer
-
One-Step Generative Modeling via Wasserstein Gradient Flows
W-Flow compresses a Wasserstein gradient flow defined via Sinkhorn divergence into a single-step neural generator, reporting 1.29 FID on ImageNet 256x256 with improved mode coverage.
-
Drifting Field Policy: A One-Step Generative Policy via Wasserstein Gradient Flow
DFP is a one-step generative policy using Wasserstein gradient flow on a drifting model backbone, with a top-K behavior cloning surrogate, that reaches SOTA on Robomimic and OGBench manipulation tasks.
-
Identifiability and Stability of Generative Drifting with Companion-Elliptic Kernel Families
The paper proves identifiability of drifting fields for companion-elliptic kernels and shows that field convergence plus a C0 observable recovers weak convergence even without tightness.