DriftXpress: Faster Drifting Models via Projected RKHS Fields

• URL: http://arxiv.org/abs/2605.12183v1
• Date: Tue, 12 May 2026 14:26:31 GMT
• Title: DriftXpress: Faster Drifting Models via Projected RKHS Fields
• Abstract summary: DriftXpress is an accelerated formulation of drifting models based on RKHS fields.<n>It preserves the attraction-repulsion structure of the original drifting field while reducing the cost of field evaluation.<n>Across image-generation benchmarks, DriftXpress achieves comparable FID to standard drifting while reducing wall-clock training cost.
• Score: 7.80855192041271
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Drifting Models have emerged as a new paradigm for one-step generative modeling, achieving strong image quality without iterative inference. The premise is to replace the iterative denoising process in diffusion models with a single evaluation of a generator. However, this creates a different trade-off: drifting reduces inference cost by moving much of the computation into training. We introduce DriftXpress, an accelerated formulation of drifting models based on projected RKHS fields. DriftXpress approximates the drifting kernel in a low-rank feature space. This preserves the attraction-repulsion structure of the original drifting field while reducing the cost of field evaluation. Across image-generation benchmarks, DriftXpress achieves comparable FID to standard drifting while reducing wall-clock training cost. These results show that the training-inference trade-off of drifting models can be pushed further without giving up their one-step inference advantage.

Related papers

• Scaling Reinforcement Learning for Diffusion Models via Velocity Matching [66.34541945134916]
  RewardBench fine-tuning is an important tool for adapting diffusion models to human preferences and task-specific objectives.<n>Unlike autoregressive models, diffusion models do not provide tractable likelihoods for generated samples.<n>We propose reward-based velocity matching (RVM), a simple trajectory-free update that acts directly on the velocity field.
  arXiv  Detail & Related papers  (2026-08-24T17:13:43Z)
• Second Order Drifting Models [9.446439690148713]
  Drifting models evolve the model distribution during training using a predefined sample-based drift field.<n>We propose Second-Order Drifting Models, which lift drifting dynamics into phase space by augmenting generated samples with artificial velocity variables.<n>We derive a practical semi-implicit training algorithm and evaluate it on synthetic distribution matching, sequential data generation, and robotic control.
  arXiv  Detail & Related papers  (2026-08-08T05:02:55Z)
• Perceptual Flow Matching for Few-Step Generative Modeling [64.26537634881244]
  Perceptual Flow Matching is a framework for few-step generation in flow-matching models.<n>We show that PFM produces high-quality few-step generators when supervised in an appropriate representation space.
  arXiv  Detail & Related papers  (2026-07-03T17:55:01Z)
• Distilling Drifting Transformers with Representation Autoencoders [23.990618033500695]
  Representation Autoencoders (RAEs) have improved diffusion and flow models by semantically richer latent space.<n>In this work, we argue that the RAE latent space is compatible with distillation via the newly proposed Drifting Models.
  arXiv  Detail & Related papers  (2026-06-14T02:46:50Z)
• Drift Flow Matching [69.16550765154163]
  We propose Drift Flow Matching, a framework that connects drifting generative modeling with flow-based iterative generation.<n>DFM preserves the efficiency of direct transport maps while enabling generation to be refined through multiple inference steps.
  arXiv  Detail & Related papers  (2026-05-17T03:58:01Z)
• Kernel-Gradient Drifting Models [34.79469799316397]
  kernel-gradient drifting is a one-step generative modeling framework.<n>It replaces the fixed Euclidean displacement direction in drifting models with directions induced by the kernel itself.
  arXiv  Detail & Related papers  (2026-05-11T15:33:21Z)
• Teacher-Feature Drifting: One-Step Diffusion Distillation with Pretrained Diffusion Representations [48.0169045294604]
  We show that a single drifting loss can be directly used to simplify one step distillation.<n>Unlike the original Drifting Model, which relies on an additional pretrained feature extractor, we use intermediate hidden states of the pretrained teacher model as the feature representation.<n>Our method achieves efficient one step generation with competitive image quality and diversity, achieving FID scores of 1.58 on ImageNet-64$times$64 and 18.4 on SDXL.
  arXiv  Detail & Related papers  (2026-05-08T06:33:33Z)
• Lookahead Drifting Model [20.728314672584375]
  A new paradigm named emphdrifting model has been proposed for mapping distributions, which achieves the SOTA image generation performance over ImageNet via one-step neural functional evaluation (NFE)<n>The basic idea is to compute a drifting term at each training iteration and then push the output of the model towards the direction of the drifting term.<n> Experimental results on toy examples and CIFAR10 demonstrate the better performance of the new method than the baseline.
  arXiv  Detail & Related papers  (2026-04-10T11:12:21Z)
• BAgger: Backwards Aggregation for Mitigating Drift in Autoregressive Video Diffusion Models [50.986189632485285]
  We introduce Backwards Aggregation (BAgger), a self-supervised scheme that constructs corrective trajectories from the model's own rollouts.<n>Unlike prior approaches that rely on few-step distillation and distribution-matching losses, BAgger trains with standard score or flow matching objectives.<n>We instantiate BAgger on causal diffusion transformers and evaluate on text-to-video, video extension, and multi-prompt generation.
  arXiv  Detail & Related papers  (2025-12-12T23:02:02Z)
• DriftLite: Lightweight Drift Control for Inference-Time Scaling of Diffusion Models [22.823183347642132]
  We study inference-time scaling for diffusion models, where the goal is to adapt a pre-trained model to new target distributions without retraining.<n>We introduce DriftLite, a lightweight, training-free particle-based approach that steers the inference dynamics on the fly with provably optimal stability.
  arXiv  Detail & Related papers  (2025-09-25T22:21:59Z)
• One Diffusion Step to Real-World Super-Resolution via Flow Trajectory Distillation [60.54811860967658]
  FluxSR is a novel one-step diffusion Real-ISR based on flow matching models.<n>First, we introduce Flow Trajectory Distillation (FTD) to distill a multi-step flow matching model into a one-step Real-ISR.<n>Second, to improve image realism and address high-frequency artifact issues in generated images, we propose TV-LPIPS as a perceptual loss.
  arXiv  Detail & Related papers  (2025-02-04T04:11:29Z)
• DriftGAN: Using historical data for Unsupervised Recurring Drift Detection [0.6358693097475243]
  In real-world applications, input data distributions are rarely static over a period of time, a phenomenon known as concept drift. Most concept drift detection methods work on detecting concept drifts and signalling the requirement to retrain the model. We present an unsupervised method based on Generative Adversarial Networks(GAN) to detect concept drifts and identify whether a specific concept drift occurred in the past.
  arXiv  Detail & Related papers  (2024-07-09T04:38:44Z)
• ReDi: Efficient Learning-Free Diffusion Inference via Trajectory Retrieval [68.7008281316644]
  ReDi is a learning-free Retrieval-based Diffusion sampling framework. We show that ReDi improves the model inference efficiency by 2x speedup.
  arXiv  Detail & Related papers  (2023-02-05T03:01:28Z)
• Minimizing Trajectory Curvature of ODE-based Generative Models [45.89620603363946]
  Recent generative models, such as diffusion models, rectified flows, and flow matching, define a generative process as a time reversal of a fixed forward process. We present an efficient method of training the forward process to minimize the curvature of generative trajectories without any ODE/SDE simulation.
  arXiv  Detail & Related papers  (2023-01-27T21:52:03Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.