WISE: World-model-guided Imagination Scheduling for Efficient Post-training of Vision-Language-Action Models

• URL: http://arxiv.org/abs/2609.03681v1
• Date: Thu, 03 Sep 2026 11:17:57 GMT
• Title: WISE: World-model-guided Imagination Scheduling for Efficient Post-training of Vision-Language-Action Models
• Abstract summary: Post-training VLA policies typically rely on supervised fine-tuning with expensive and potentially unstable real-world exploration.<n>We introduce WISE (World-model-guided Imagination Scheduling for Efficient Post-training of Vision-Language-Action Models), a unified framework that coordinates when and how world-model imagination is used during policy refinement.
• Score: 7.861070986808876
• License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/
• Abstract: Post-training VLA policies typically rely on supervised fine-tuning with costly expert demonstrations or reinforcement learning with expensive and potentially unstable real-world exploration. World models offer a promising alternative by evaluating candidate behaviors through imagined futures, yet effective post-training requires more than accurate prediction: imagination must be scheduled where it is useful, bounded within reliable horizons, and translated into trustworthy policy supervision. In robotic manipulation, the value of imagination varies substantially across execution stages, while extended rollouts can accumulate prediction errors and introduce unreliable learning signals. We introduce WISE (World-model-guided Imagination Scheduling for Efficient Post-training of Vision-Language-Action Models), a unified framework that coordinates when and how world-model imagination is used during policy refinement. WISE selectively invokes imagination at interaction-relevant states, performs bounded multi-view rollouts, evaluates candidate futures using progress and completion signals, and uses their relative outcomes to refine actions generated from real interaction contexts. Extensive experiments with both $π_0$ and $π_{0.5}$ demonstrate consistent improvements across diverse manipulation tasks while reducing GPU computation time by approximately 80% compared with full imagination. Real-world evaluations further show substantial gains in robustness and generalization under diverse real-world distribution shifts.

Related papers

• SyncWorld: Visual Calibration Enables World Models as Zero-Shot Simulators [61.95179776027961]
  SyncWorld is an action-conditioned world model that serves as a zero-shot simulator across unseen environments without any additional training.<n>Training with visual calibration contexts teaches the model to interpret actions through visual evidence.<n>Experiments show that SyncWorld can accurately simulate action outcomes in previously unseen settings.
  arXiv  Detail & Related papers  (2026-09-08T17:59:47Z)
• Do Robotic World Models Really Follow Actions? Diagnosing and Aligning Action-Conditioned Generation for Policy Learning [53.77164209208635]
  We introduce WorldEcho, which probes action following over a broader action distribution using visual integrity and trajectory alignment.<n>We propose WorldSync, which strengthens action following along three axes: distributional coverage, representational grounding, and intervention-effect alignment.<n>Experiments on RoboTwin benchmarks and real-robot tasks show that WorldSync improves WorldEcho metrics and serves as a more reliable simulator for iterative policy improvement.
  arXiv  Detail & Related papers  (2026-08-25T17:59:49Z)
• Predicting Consequences and Reinforcing Navigation Policies with Latent World Models [31.865100130203317]
  We propose a compatibility prediction Latent World Model (LWM) for robot navigation.<n>Our key insight is that spatial proximity correlates with latent feature similarity, enabling action consequences to be evaluated directly in latent space.<n>Our approach significantly outperforms prior world model and imitation learning methods in prediction accuracy, policy learning, and real-world navigation performance.
  arXiv  Detail & Related papers  (2026-08-23T11:58:42Z)
• World Models Meet Language Models: On the Complementarity of Concrete and Abstract Reasoning [67.93052893044603]
  We study how a model learns to invoke, verify, and integrate visual future simulation alongside abstract reasoning.<n>PF-OPSD uses ground-truth future videos and answers only as teacher-side privileged context to evaluate on-policy concrete-reasoning trajectories.
  arXiv  Detail & Related papers  (2026-06-02T13:07:49Z)
• Learning Vision-Language-Action World Models for Autonomous Driving [15.103497388527943]
  Vision-Language-Action (VLA) models have recently achieved notable progress in end-to-end autonomous driving.<n>VLA models often lack explicit modeling of temporal dynamics and global world consistency.<n>We present VLA-World, a simple yet effective VLA world model that unifies predictive imagination with reflective reasoning.
  arXiv  Detail & Related papers  (2026-04-10T07:38:05Z)
• LatentPilot: Scene-Aware Vision-and-Language Navigation by Dreaming Ahead with Latent Visual Reasoning [51.969318585152116]
  LatentPilot exploits future observations during training as a valuable data source to learn action-conditioned visual dynamics.<n>We propose a flywheel-style training mechanism that iteratively collects on-policy trajectories and retrains the model to better match the agent's behavior distribution.<n> Experiments on R2R-CE, RxR-CE, and R2R-PE benchmarks achieve new SOTA results, and real-robot tests across diverse environments demonstrate LatentPilot's superior understanding of environment-action dynamics in scene.
  arXiv  Detail & Related papers  (2026-03-31T02:21:59Z)
• Self-Correcting VLA: Online Action Refinement via Sparse World Imagination [55.982504915794514]
  We propose Self-Correcting VLA (SC-VLA), which achieve self-improvement by intrinsically guiding action refinement through sparse imagination.<n>SC-VLA achieve state-of-the-art performance, yielding the highest task throughput with 16% fewer steps and a 9% higher success rate than the best-performing baselines.
  arXiv  Detail & Related papers  (2026-02-25T06:58:06Z)
• World-VLA-Loop: Closed-Loop Learning of Video World Model and VLA Policy [55.03832008486675]
  World-VLA-Loop is a closed-loop framework for the joint refinement of world models and Vision-Language-Action (VLA) policies.<n>We propose a state-aware video world model that functions as a high-fidelity interactive simulator by jointly predicting future observations and reward signals.
  arXiv  Detail & Related papers  (2026-02-06T08:57:55Z)
• AstraNav-World: World Model for Foresight Control and Consistency [40.07910402326578]
  Embodied navigation in dynamic environments demands accurate foresight of how the world will evolve and how actions will unfold over time.<n>We propose AstraNav-World, an end-to-end world model that jointly reasons about future visual states and action sequences.<n>Our framework integrates a diffusion-based video generator with a vision-language policy, enabling synchronized rollouts.
  arXiv  Detail & Related papers  (2025-12-25T15:31:24Z)
• WorldGym: World Model as An Environment for Policy Evaluation [41.204900701616914]
  WorldGym is an autoregressive, action-conditioned video generation model which serves as a proxy to real world environments.<n> Policies are evaluated via Monte Carlo rollouts in the world model, with a vision-language model providing rewards.<n>We show that WorldGym is able to preserve relative policy rankings across different policy versions, sizes, and training checkpoints.
  arXiv  Detail & Related papers  (2025-05-31T15:51:56Z)

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.