X-Cache: Cross-Chunk Block Caching for Few-Step Autoregressive World Models Inference
Fuente:
arXiv
Salvato in:
| Autori principali: | , , , , , , , , , , , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2026
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866915983286861824 |
|---|---|
| author | Zeng, Yixiao Zheng, Jianlei Zheng, Chaoda Chen, Shijia Liu, Mingdian Liu, Tongping Luo, Tengwei Zhang, Yu Wang, Boyang Xu, Linkun Lu, Siyuan Tian, Bo Liu, Xianming |
| author_facet | Zeng, Yixiao Zheng, Jianlei Zheng, Chaoda Chen, Shijia Liu, Mingdian Liu, Tongping Luo, Tengwei Zhang, Yu Wang, Boyang Xu, Linkun Lu, Siyuan Tian, Bo Liu, Xianming |
| contents | Real-time world simulation is becoming a key infrastructure for scalable evaluation and online reinforcement learning of autonomous driving systems. Recent driving world models built on autoregressive video diffusion achieve high-fidelity, controllable multi-camera generation, but their inference cost remains a bottleneck for interactive deployment. However, existing diffusion caching methods are designed for offline video generation with multiple denoising steps, and do not transfer to this scenario. Few-step distilled models have no inter-step redundancy left for these methods to reuse, and sequence-level parallelization techniques require future conditioning that closed-loop interactive generation does not provide. We present X-Cache, a training-free acceleration method that caches along a different axis: across consecutive generation chunks rather than across denoising steps. X-Cache maintains per-block residual caches that persist across chunks, and applies a dual-metric gating mechanism over a structure- and action-aware block-input fingerprint to independently decide whether each block should recompute or reuse its cached residual. To prevent approximation errors from permanently contaminating the autoregressive KV cache, X-Cache identifies KV update chunks (the forward passes that write clean keys and values into the persistent cache) and unconditionally forces full computation on these chunks, cutting off error propagation. We implement X-Cache on X-world, a production multi-camera action-conditioned driving world model built on multi-block causal DiT with few-step denoising and rolling KV cache. X-Cache achieves 71% block skip rate with 2.6x wall-clock speedup while maintaining minimum degradation. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2604_20289 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | X-Cache: Cross-Chunk Block Caching for Few-Step Autoregressive World Models Inference Zeng, Yixiao Zheng, Jianlei Zheng, Chaoda Chen, Shijia Liu, Mingdian Liu, Tongping Luo, Tengwei Zhang, Yu Wang, Boyang Xu, Linkun Lu, Siyuan Tian, Bo Liu, Xianming Computer Vision and Pattern Recognition Real-time world simulation is becoming a key infrastructure for scalable evaluation and online reinforcement learning of autonomous driving systems. Recent driving world models built on autoregressive video diffusion achieve high-fidelity, controllable multi-camera generation, but their inference cost remains a bottleneck for interactive deployment. However, existing diffusion caching methods are designed for offline video generation with multiple denoising steps, and do not transfer to this scenario. Few-step distilled models have no inter-step redundancy left for these methods to reuse, and sequence-level parallelization techniques require future conditioning that closed-loop interactive generation does not provide. We present X-Cache, a training-free acceleration method that caches along a different axis: across consecutive generation chunks rather than across denoising steps. X-Cache maintains per-block residual caches that persist across chunks, and applies a dual-metric gating mechanism over a structure- and action-aware block-input fingerprint to independently decide whether each block should recompute or reuse its cached residual. To prevent approximation errors from permanently contaminating the autoregressive KV cache, X-Cache identifies KV update chunks (the forward passes that write clean keys and values into the persistent cache) and unconditionally forces full computation on these chunks, cutting off error propagation. We implement X-Cache on X-world, a production multi-camera action-conditioned driving world model built on multi-block causal DiT with few-step denoising and rolling KV cache. X-Cache achieves 71% block skip rate with 2.6x wall-clock speedup while maintaining minimum degradation. |
| title | X-Cache: Cross-Chunk Block Caching for Few-Step Autoregressive World Models Inference |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2604.20289 |