RISE: Adaptive Imagination for World Action Models
Summary
RISE introduces an adaptive framework for imagination rollouts in world action models, using a counterfactual driving dataset to improve planning performance while reducing unnecessary computation.
View Cached Full Text
Cached at: 08/25/26, 04:34 AM
Paper page - RISE: Adaptive Imagination for World Action Models
Source: https://huggingface.co/papers/2608.20430
Abstract
RISE adaptively decides when to continue or stop imagination rollouts for planning by weighing expected benefit against cost, supported by a counterfactual driving dataset with expert annotations.
World Action Models(WAMs) improve planning by incorporating future world evolution into action generation, yet existing methods allocate a fixed imagination budget to every scene. We propose RISE (Refining Imagination through SElective Rollout), a system-leveladaptive imaginationframework that makes sequentialRoll/Stop decisionsaccording to the expected planning benefit of continued rollout. At each step, aLatent Evaluatorestimates the risk revealed by the current prefix and how much planning could improve if imagination continues, while aRollout Gateweighs this expected benefit against additional computation cost. Since factual driving logs expose only one realized future, we further constructCounterDrive, acounterfactual datasetwith diverse outcomes and risk levels, to enrich future dynamics and provide localizedrisk supervision. Each retained sample undergoes expert verification and annotation of trajectory validity, incident onset, and causal category, providing a reusable resource for safety-critical world-modeling research. Experiments onNAVSIMandnuScenesshow that RISE achieves the best overall planning performance while reducing unnecessary rollout, with additional transfer results supporting its plug-in generality across WAM architectures.
View arXiv pageView PDFProject pageGitHub20Add to collection
Get this paper in your agent:
hf papers read 2608\.20430
Don’t have the latest CLI?curl \-LsSf https://hf\.co/cli/install\.sh \| bash
Models citing this paper0
No model linking this paper
Cite arxiv.org/abs/2608.20430 in a model README.md to link it from this page.
Datasets citing this paper0
No dataset linking this paper
Cite arxiv.org/abs/2608.20430 in a dataset README.md to link it from this page.
Spaces citing this paper0
No Space linking this paper
Cite arxiv.org/abs/2608.20430 in a Space README.md to link it from this page.
Collections including this paper0
No Collection including this paper
Add this paper to acollectionto link it from this page.
Similar Articles
The DAWN of World-Action Interactive Models
This paper introduces DAWN, a latent generative baseline for World-Action Interactive Models (WAIMs) that jointly models scene evolution and action generation through recursive refinement, achieving strong long-horizon planning in autonomous driving scenarios.
Second Thoughts: Exploring and Planning through Comparative Imagination in World Models
The paper introduces SDI (Sibling Disagreement in Imagination), a method that unifies exploration and planning in world-model-based RL by using the actor's own rollout variability as a shared reference to decide when a planned action should replace the learned policy, outperforming baselines on Meta-World, Adroit, and DM-Control.
Rolling-WAM: World Action Models with Rolling Imagination
Rolling-WAM introduces a method to distribute denoising across replanning cycles in world action models for robotic manipulation, achieving a 4.5x speedup in replanning while maintaining competitive performance.
When to Trust Imagination: Adaptive Action Execution for World Action Models
This paper introduces FFDC, a lightweight verifier for World Action Models that enables adaptive action chunk sizes by checking consistency between predicted and actual observations, improving efficiency and robustness in robotic manipulation.
OSWAM: An Effective World Action Model with One-Step Action Generation
OSWAM introduces a world action model that enables one-step diffusion-based action generation for autonomous driving and embodied control by initializing with imagination tokens and distilling future-conditioned privileged information, achieving a ~3x speedup and 40ms latency on an RTX 5090 while matching multi-step baselines.