Drift-AR：通过反向对称漂移实现单步视觉自回归生成
Drift-AR: Single-Step Visual Autoregressive Generation via Anti-Symmetric Drifting
AI总结：
Drift-AR通过利用熵信号加速自回归和视觉解码阶段，实现单步生成，提升了生成速度并保持了高质量。
摘要：
自回归（AR）-扩散混合范式结合了AR的结构化语义建模与扩散的高保真合成，但存在双重速度瓶颈：顺序AR阶段和迭代多步去噪的扩散视觉解码阶段。现有方法各自解决而无统一设计。我们发现连续空间AR模型的每个位置预测熵自然编码空间变化的生成不确定性，同时控制AR阶段的草稿预测质量和反映视觉解码阶段所需的修正努力。由于熵与两个瓶颈内在相关，它成为联合加速的自然统一信号。本文提出Drift-AR，利用熵信号加速两个阶段：1）对于AR加速，引入基于熵的推测解码，通过因果归一化熵损失调整草稿-目标熵分布，解决导致过度草稿拒绝的熵不匹配；2）对于视觉解码加速，将熵重新解释为初始状态的物理方差，用于反向对称漂移场——高熵位置更强地漂向数据流形，而低熵位置漂移消失——实现单步（1-NFE）解码，无需迭代去噪或蒸馏。此外，两个阶段共享相同的熵信号，计算一次无额外成本。在MAR、TransDiff和NextStep-1上的实验显示，速度提升3.8-5.5倍，具有真实的1-NFE解码，匹配或超越原始质量。代码将在https://github.com/aSleepyTree/Drift-AR上提供。
英文摘要：
Autoregressive (AR)-Diffusion hybrid paradigms combine AR's structured semantic modeling with diffusion's high-fidelity synthesis, yet suffer from a dual speed bottleneck: the sequential AR stage and the iterative multi-step denoising of the diffusion vision decode stage. Existing methods address each in isolation without a unified principle design. We observe that the per-position \emph{prediction entropy} of continuous-space AR models naturally encodes spatially varying generation uncertainty, which simultaneously governing draft prediction quality in the AR stage and reflecting the corrective effort required by vision decoding stage, which is not fully explored before. Since entropy is inherently tied to both bottlenecks, it serves as a natural unifying signal for joint acceleration. In this work, we propose \textbf{Drift-AR}, which leverages entropy signal to accelerate both stages: 1) for AR acceleration, we introduce Entropy-Informed Speculative Decoding that align draft-target entropy distributions via a causal-normalized entropy loss, resolving the entropy mismatch that causes excessive draft rejection; 2) for visual decoder acceleration, we reinterpret entropy as the \emph{physical variance} of the initial state for an anti-symmetric drifting field -- high-entropy positions activate stronger drift toward the data manifold while low-entropy positions yield vanishing drift -- enabling single-step (1-NFE) decoding without iterative denoising or distillation. Moreover, both stages share the same entropy signal, which is computed once with no extra cost. Experiments on MAR, TransDiff, and NextStep-1 demonstrate 3.8-5.5$\times$ speedup with genuine 1-NFE decoding, matching or surpassing original quality. Code will be available at https://github.com/aSleepyTree/Drift-AR.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
