用自监督回溯机制，让视频生成模型学会纠正自己犯的错。
BAgger: Backwards Aggregation for Mitigating Drift in Autoregressive Video Diffusion Models
- 通过模型自身生成的轨迹构建纠错路径，反向训练恢复能力。
- 在长序列生成中运动更稳定，视觉一致性提升，漂移现象减少。
- 无需大模型教师或复杂损失，适合实际部署的视频生成任务。
自回归视频模型通过预测下一帧展现世界建模潜力，但存在暴露偏差：训练时使用干净上下文，推理时依赖自生成帧，导致错误累积与质量漂移。本文提出后向聚合（BAgger），一种自监督方案，利用模型自身滚动生成的轨迹构建纠正路径，教会模型从错误中恢复。与依赖少步蒸馏和分布匹配损失的先前方法不同，BAgger采用标准得分匹配或流匹配目标进行训练，避免使用大型教师模型和长时间反向传播。我们在因果扩散变换器上实现BAgger，评估于文本到视频、视频扩展和多提示生成任务，观察到更稳定的长时序运动表现和更好的视觉一致性，显著降低了漂移现象。
原文摘要 · Abstract (English)
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 errors to compound and quality to drift over time. We introduce Backwards Aggregation (BAgger), a self-supervised scheme that constructs corrective trajectories from the model's own rollouts, teaching it to recover from its mistakes. Unlike prior approaches that rely on few-step distillation and distribution-matching losses, which can hurt quality and diversity, BAgger trains with standard score or flow matching objectives, avoiding large teachers and long-chain backpropagation through time. We instantiate BAgger on causal diffusion transformers and evaluate on text-to-video, video extension, and multi-prompt generation, observing more stable long-horizon motion and better visual consistency with reduced drift.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。