DISK: Dynamic Inference SKipping for World Models
Fuente:
arXiv
Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866911412424540160 |
|---|---|
| author | Naman, Anugunj Zhang, Gaibo Singh, Ayushman Zhang, Yaguang |
| author_facet | Naman, Anugunj Zhang, Gaibo Singh, Ayushman Zhang, Yaguang |
| contents | We present DISK, a training-free adaptive inference method for autoregressive world models. DISK coordinates two coupled diffusion transformers for video and ego-trajectory via dual-branch controllers with cross-modal skip decisions, preserving motion-appearance consistency without retraining. We extend higher-order latent-difference skip testing to the autoregressive chain-of-forward regime and propagate controller statistics through rollout loops for long-horizon stability. When integrated into closed-loop driving rollouts on 1500 NuPlan and NuScenes samples using an NVIDIA L40S GPU, DISK achieves 2x speedup on trajectory diffusion and 1.6x speedup on video diffusion while maintaining L2 planning error, visual quality (FID/FVD), and NAVSIM PDMS scores, demonstrating practical long-horizon video-and-trajectory prediction at substantially reduced cost. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2602_00440 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | DISK: Dynamic Inference SKipping for World Models Naman, Anugunj Zhang, Gaibo Singh, Ayushman Zhang, Yaguang Computer Vision and Pattern Recognition Machine Learning Robotics We present DISK, a training-free adaptive inference method for autoregressive world models. DISK coordinates two coupled diffusion transformers for video and ego-trajectory via dual-branch controllers with cross-modal skip decisions, preserving motion-appearance consistency without retraining. We extend higher-order latent-difference skip testing to the autoregressive chain-of-forward regime and propagate controller statistics through rollout loops for long-horizon stability. When integrated into closed-loop driving rollouts on 1500 NuPlan and NuScenes samples using an NVIDIA L40S GPU, DISK achieves 2x speedup on trajectory diffusion and 1.6x speedup on video diffusion while maintaining L2 planning error, visual quality (FID/FVD), and NAVSIM PDMS scores, demonstrating practical long-horizon video-and-trajectory prediction at substantially reduced cost. |
| title | DISK: Dynamic Inference SKipping for World Models |
| topic | Computer Vision and Pattern Recognition Machine Learning Robotics |
| url | https://arxiv.org/abs/2602.00440 |