AnyStep-WAM: Budget-Aligned Distillation and Adaptive Inference for World Action Models

arXiv:2609.33748 · cs.RO · Submitted 2026-09-27 · Read on arXiv

cs.RO

Submitted: 2026-09-27

Updated: 2026-09-30

Terminology

• Distillation — 26× in this paper · explained in 2 episodes
• Fidelity — 24× in this paper · explained in 2 episodes
• Ablation — 10× in this paper · explained in 1 episode
• Flow Matching — 9× in this paper · explained in 8 episodes
• Adaptation — 5× in this paper · explained in 1 episode
• Consistency — 5× in this paper · explained in 3 episodes
• Distribution Matching — 4× in this paper · explained in 2 episodes
• LoRA Adapters — 4× in this paper · explained in 3 episodes
• Transformer — 4× in this paper · explained in 2 episodes
• Adaptive Scheduler — 3× in this paper · explained in 1 episode
• Conference on Computer Vision and Pattern Recognition — 3× in this paper
• Generalization — 3× in this paper · explained in 13 episodes
• Benchmark — 2× in this paper · explained in 2 episodes
• Diffusion Model — 2× in this paper · explained in 8 episodes
• Generative Model — 2× in this paper · explained in 2 episodes
• Generative Modeling — 2× in this paper · explained in 3 episodes
• Generative Models — 2× in this paper · explained in 2 episodes
• Reinforcement Learning — 2× in this paper · explained in 4 episodes
• Accuracy — 1× in this paper · explained in 1 episode
• AdaVLA — 1× in this paper · explained in 1 episode
• Consistency Distillation — 1× in this paper · explained in 2 episodes
• Cosine Similarity — 1× in this paper · explained in 3 episodes
• Diffusion Models — 1× in this paper · explained in 6 episodes
• Embeddings — 1× in this paper · explained in 2 episodes
• Generative AI — 1× in this paper · explained in 2 episodes
• VLA Models — 1× in this paper · explained in 3 episodes
• World Model — 1× in this paper · explained in 7 episodes

Sources

• Flash-WAM: Modality-Aware Distillation for World Action Models
• Adaptive-WAM: Quality-Guided Early-Exit Planning from Intermediate Video-Diffusion Features
• RoboTwin 2.0: A Scalable Data Generator and Benchmark with Strong Domain Randomization for Robust Bimanual Robotic Manipulation
• Let It Be Simple: One-Step Action Generation for Vision-Language-Action Models
• Improved Mean Flows: On the Challenges of Fastforward Generative Models
• AdaVLA: Adaptive Step Flow Matching for Training-free Acceleration of Vision-Language-Action Models
• LoRA: Low-Rank Adaptation of Large Language Models
• DSA: Dynamic Step Allocation for Fast Autoregressive Video Generation
• ELASTIC: Efficiently Learning to Adaptively Scale Test-Time Compute for Generative Control Policies
• Causal World Modeling for Robot Control
• Flow Matching for Generative Modeling
• Latent Consistency Models: Synthesizing High-Resolution Images with Few-Step Inference
• The Geometry of Flow-Matching Uncertainty: A Cost-free Uncertainty Proxy and Its Application in Flow-based VLA Failure Detection
• Elastic Queries Reinforcement Learning: Self-Aware Policy Execution for VLA Models
• ManiFlow: A General Robot Manipulation Policy via Consistency Flow Training
• Stable Velocity: A Variance Perspective on Flow Matching
• D3P: Dynamic Denoising Diffusion Policy via Reinforcement Learning
• Fast-WAM: Do World Action Models Need Test-time Future Imagination?

Related papers

• FMT x: An Efficient and Asymptotically Optimal Extension of the Fast Marching Tree for Dynamic Replanning
• MPCFormer: A physics-informed data-driven approach for explainable socially-aware autonomous driving
• RoboLab: A High-Fidelity Simulation Benchmark for Analysis of Task Generalist Policies
• HRDexDB: A 4D Dexterous Grasping Dataset Across Human and Multiple Robot Embodiments
• APT: Action Expert Pretraining Improves Instruction Generalization of Vision-Language-Action Policies
• Fine-tuning is Not Enough: A Parallel Framework for Collaborative Imitation and Reinforcement Learning in End-to-end Autonomous Driving