paper-with-me

홈 › Papers

HorizonDrive: Self-Corrective Autoregressive World Model for Long-horizon Driving Simulation

2026-05-12 · Conglang Zhang, Yifan Zhan, Qingjie Wang, Zhanpeng Ouyang, Yu Li, Zihao Yang, Xiaoyang Guo, Weiqiang Ren, Qian Zhang, Zhen Dong, Yinqiang Zheng, Wei Yin, Zhengqing Chen arxiv

Closed-loop driving simulation requires real-time interaction beyond short offline clips, pushing current driving world models toward autoregressive (AR) rollout. Existing AR distillation approaches typically rely on frame sinks or student-side degradation training. The former transfers poorly to driving due to fast ego-motion and rapid scene changes, while the latter remains bounded by the teacher's single-pass output length and thus provides only a limited supervision horizon. A natural question is: can the teacher itself be extended via AR rollout to provide unbounded-horizon supervision at bounded memory cost? The key difficulty is that a standard teacher drifts under its own predictions, contaminating the supervision it provides. Our key insight is to make the teacher rollout-capable, ensuring reliable supervision from its own AR rollouts. This is instantiated as HorizonDrive, an anti-drifting training-and-distillation framework for AR driving simulation. First, scheduled rollout recovery (SRR) trains the base model to reconstruct ground-truth future clips from prediction-corrupted histories, yielding a teacher that remains stable across long AR rollouts. Second, the rollout-capable teacher is extended via AR rollout, providing long-horizon distribution-matching supervision under bounded memory, while a short-window student aligns to it with teacher rollout DMD (TRD) for efficient real-time deployment. HorizonDrive natively supports minute-scale AR rollout under bounded memory; on nuScenes, HorizonDrive reduces FID by 52% and FVD by 37%, and lowers ARE and DTW by 21% and 9% relative to the strongest long-horizon streaming baselines, while remaining competitive with single-pass driving video generators.

📄 PDF Abstract BibTeX arXiv:2605.11596

Code (0)

등록된 구현이 없습니다.

Similar Papers 제목 키워드 기반

BAgger: Backwards Aggregation for Mitigating Drift in Autoregressive Video Diffusion Models

2025-12-12 · Ryan Po, Eric Ryan Chan, Changan Chen, Gordon Wetzstein arxiv

Autoregressive video models are promising for world modeling via next-frame prediction, but they suffer from exposure bias: a mismatch between training on clean contexts and inference on self-generated frames, causing er…

OPSD-V: On-Policy Self-Distillation for Post-Training Few-Step Autoregressive Video Generators

2026-07-09 · Hongyu Liu, Chun Wang, Feng Gao, Xuanhua He 외 arxiv

We propose OPSD-V, an on-policy self-distillation paradigm for post-training few-step autoregressive (AR) video diffusion models. Existing few-step AR video generators can produce long videos with low latency, but still …

McCast: Memory-Guided Latent Drift Correction for Long-Horizon Precipitation Nowcasting

2026-05-13 · Penghui Wen, Yu Luo, Lintao Wang, Mengwei He 외 arxiv

Existing precipitation nowcasting methods typically adopt an autoregressive formulation, where future states are predicted from previous outputs. However, such an approach accumulates errors over long rollouts, causing f…

Self-Corrective Task Planning by Inverse Prompting with Large Language Models

2025-03-10 · Jiho Lee, Hayun Lee, Jonghyeon Kim, Kyungjae Lee 외

In robot task planning, large language models (LLMs) have shown significant promise in generating complex and long-horizon action sequences. However, it is observed that LLMs often produce responses that sound plausible …

Robot Task PlanningTask Planning

Self-Reflective Generation at Test Time

2025-10-03 · Jian Mu, Qixin Zhang, Zhiyong Wang, Menglin Yang 외 arxiv

Large language models (LLMs) increasingly solve complex reasoning tasks via long chain-of-thought, but their forward-only autoregressive generation process is fragile; early token errors can cascade, which creates a clea…

Mathematical Reasoning