Action-Inspired Generative Models

• URL: http://arxiv.org/abs/2605.14631v1
• Date: Thu, 14 May 2026 09:43:32 GMT
• Title: Action-Inspired Generative Models
• Abstract summary: Action-Inspired Generative Models (AGMs)<n>We introduce a dual-network generative framework motivated by the observation that existing bridge-matching methods assign uniform regression weight to every transition in the transport landscape.<n>We demonstrate that selectively penalising uninformative transport paths through the learned potential yields consistent improvements in generation quality across fidelity and coverage metrics.
• Score: 0.0
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: We introduce Action-Inspired Generative Models (AGMs), a dual-network generative framework motivated by the observation that existing bridge-matching methods assign uniform regression weight to every stochastic transition in the transport landscape, regardless of whether a given bridge sample lies along a structurally coherent trajectory or a degenerate one. We address this by introducing a lightweight learned scalar potential $V_φ$ that scores bridge samples online and modulates the drift objective via importance weights derived through a stop-gradient barrier -- preventing adversarial feedback between the two networks whilst preserving $V_φ$'s guiding signal. Crucially, $V_φ$ comprises only $\sim$1.4% of the primary drift network's parameter count, adds no overhead to the inference graph, and requires no iterative half-bridge fitting or auxiliary stochastic differential equation (SDE) solvers: it is a plug-and-play enhancement to any bridge-matching training loop. At inference, $V_φ$ is discarded entirely, leaving standard Euler-Maruyama integration of the exponential moving average (EMA) drift. We demonstrate that selectively penalising uninformative transport paths through the learned potential yields consistent improvements in generation quality across fidelity and coverage metrics.

Related papers

• Sparse Weight Decomposition for Efficient Circuit Extraction [36.87892448834478]
  Dense pretrained transformers do not naturally expose interpretable units for circuit extraction.<n>We propose Sparse Weight Decomposition, which factorizes pretrained projections by factorizing each weight matrix into two sparse factors.<n>We show that SWD remains effective for full-model replacement of all attention and weight matrices after fine-tuning the nonzero factor values.
  arXiv  Detail & Related papers  (2026-08-04T16:40:48Z)
• Improving Neural Network Training by Decoupling the Magnitude and Direction of Weight Vectors [110.10775872372047]
  We propose emphMagnitude-Direction-Direction (MD) Decoupling.<n>It factorizes each weight into a fixed-norm direction on a hypersphere and learnable per-row and per-column magnitude gains, updated at separate learning rates.<n>Across both Adam and Muon, MD Decoupling improves on well-tuned baselines, transfers the optimal LR across model width without retuning, and continues to help at scale on large Mixture-of-Experts models.
  arXiv  Detail & Related papers  (2026-06-24T15:40:26Z)
• Probabilistic Recurrent Intention Switching Model [8.25071593831945]
  Inverse reinforcement learning (IRL) recovers reward functions from observed behavior, yet traditional methods assume a single stationary reward that cannot capture goal switching within an episode.<n>Recent multi-intention IRL methods address this by segmenting trajectories, but model intention transitions as either a memoryless Markov chain or via manual state augmentation with a fixed history window.<n>We propose the Probabilistic Recurrent Intention Switching Model (PRISM), which replaces both mechanisms with a lightweight recurrent network that maps observation history to a per-step intention distribution.
  arXiv  Detail & Related papers  (2026-05-26T13:19:00Z)
• Efficient Adjoint Matching for Fine-tuning Diffusion Models [47.05085426361986]
  We propose EAM, which substantially improves training efficiency by reformulating the SOC problem with a textit base drift and a textitterminal cost.<n>EAM converges up to 4x faster than AM and matches or surpasses it across various metrics including PickScore, ImageReward, HPSv2.1, CLIPScore and Aesthetics.
  arXiv  Detail & Related papers  (2026-05-12T03:55:12Z)
• Rectified Schrödinger Bridge Matching for Few-Step Visual Navigation [18.319705014361908]
  Visual navigation is a core challenge in Embodied AI, requiring autonomous agents to translate high-dimensional sensory observations into long-horizon action trajectories.<n>We propose Rectified Schrdinger Bridge Matching (RSBM), a framework that exploits a shared velocity-field structure between Schrdinger Bridges and deterministic transport.<n>RSBM achieves over 94% and 92% success rate in merely 3 stage training, substantially narrowing the gap between high-fidelity generative policies and the low-latency demands of Embodied AI.
  arXiv  Detail & Related papers  (2026-04-07T10:22:27Z)
• Look-Ahead and Look-Back Flows: Training-Free Image Generation with Trajectory Smoothing [3.77130368225397]
  Various training-free flow matching approaches have been developed to improve image generation through flow velocity field adjustment.<n>We propose two training-free trajectory smoothing schemes: emphLook-Ahead, which averages the current and next-step latents using a curvature-gated weight, and emphLook-Back, which smoothes latents using an exponential moving average with decay.
  arXiv  Detail & Related papers  (2026-02-10T06:34:47Z)
• DiffusionDriveV2: Reinforcement Learning-Constrained Truncated Diffusion Modeling in End-to-End Autonomous Driving [65.7087560656003]
  Generative diffusion models for end-to-end autonomous driving often suffer from mode collapse.<n>We propose DiffusionDriveV2, which leverages reinforcement learning to constrain low-quality modes and explore for superior trajectories.<n>This significantly enhances the overall output quality while preserving the inherent multimodality of its core Gaussian Mixture Model.
  arXiv  Detail & Related papers  (2025-12-08T17:29:52Z)
• Variational Inference Optimized Using the Curved Geometry of Coupled Free Energy [4.252549987351643]
  We introduce an optimization framework for variational inference based on the coupled free energy.<n>We extend variational inference techniques to account for the curved geometry of the coupled exponential family.
  arXiv  Detail & Related papers  (2025-06-10T10:47:21Z)
• Flow-GRPO: Training Flow Matching Models via Online RL [80.62659379624867]
  We propose Flow-GRPO, the first method to integrate online policy reinforcement learning into flow matching models.<n>Our approach uses two key strategies: (1) an ODE-to-SDE conversion that transforms a deterministic Ordinary Differential Equation into an equivalent Differential Equation (SDE) that matches the original model's marginal distribution at all timesteps; and (2) a Denoising Reduction strategy that reduces training denoising steps while retaining the original number of inference steps.
  arXiv  Detail & Related papers  (2025-05-08T17:58:45Z)
• Joint Transmit and Pinching Beamforming for Pinching Antenna Systems (PASS): Optimization-Based or Learning-Based? [89.05848771674773]
  A novel antenna system ()-enabled downlink multi-user multiple-input single-output (MISO) framework is proposed.<n>It consists of multiple waveguides, which equip numerous low-cost antennas, named (PAs)<n>The positions of PAs can be reconfigured to both spanning large-scale path and space.
  arXiv  Detail & Related papers  (2025-02-12T18:54:10Z)
• CARE Transformer: Mobile-Friendly Linear Visual Transformer via Decoupled Dual Interaction [77.8576094863446]
  We propose a new detextbfCoupled dutextbfAl-interactive lineatextbfR atttextbfEntion (CARE) mechanism. We first propose an asymmetrical feature decoupling strategy that asymmetrically decouples the learning process for local inductive bias and long-range dependencies. By adopting a decoupled learning way and fully exploiting complementarity across features, our method can achieve both high efficiency and accuracy.
  arXiv  Detail & Related papers  (2024-11-25T07:56:13Z)
• Adaptive Federated Learning Over the Air [108.62635460744109]
  We propose a federated version of adaptive gradient methods, particularly AdaGrad and Adam, within the framework of over-the-air model training. Our analysis shows that the AdaGrad-based training algorithm converges to a stationary point at the rate of $mathcalO( ln(T) / T 1 - frac1alpha ).
  arXiv  Detail & Related papers  (2024-03-11T09:10:37Z)
• Do Generative Models Know Disentanglement? Contrastive Learning is All You Need [59.033559925639075]
  We propose an unsupervised and model-agnostic method: Disentanglement via Contrast (DisCo) in the Variation Space. DisCo achieves the state-of-the-art disentanglement given pretrained non-disentangled generative models, including GAN, VAE, and Flow.
  arXiv  Detail & Related papers  (2021-02-21T08:01:20Z)

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.