Test-time scaling (TTS) has demonstrated remarkable success in enhancing large language models, yet its application to next-token prediction (NTP) autoregressive (AR) image generation remains largely uncharted. Existing TTS approaches for visual AR (VAR), which rely on frequent partial decoding and external reward models, are ill-suited for NTP-based image generation due to the inherent incompleteness of intermediate decoding results. To bridge this gap, we introduce ScalingAR, the first TTS framework specifically designed for NTP-based AR image generation that eliminates the need for early decoding or auxiliary rewards. ScalingAR leverages token entropy as a novel signal in visual token generation and operates at two complementary scaling levels: (i) Profile Level, which streams a calibrated confidence state by fusing intrinsic and conditional signals; and (ii) Policy Level, which utilizes this state to adaptively terminate low-confidence trajectories and dynamically schedule guidance for phase-appropriate conditioning strength. Experiments on both general and compositional benchmarks show that ScalingAR (1) improves base models by 12.5% on GenEval and 15.2% on TIIF-Bench, (2) efficiently reduces visual token consumption by 62.0% while outperforming baselines, and (3) successfully enhances robustness, mitigating performance drops by 26.0% in challenging scenarios.

翻译：测试时缩放（TTS）在增强大语言模型方面已展现出显著成功，但其在基于下一令牌预测（NTP）的自回归（AR）图像生成中的应用仍基本未被探索。现有的视觉自回归（VAR）TTS方法依赖于频繁的部分解码和外部奖励模型，由于中间解码结果固有的不完整性，这些方法并不适用于基于NTP的图像生成。为弥合这一差距，我们提出了ScalingAR，这是首个专门为基于NTP的AR图像生成设计的TTS框架，无需早期解码或辅助奖励。ScalingAR利用令牌熵作为视觉令牌生成中的一种新信号，并在两个互补的缩放级别上运行：（i）配置文件级别，通过融合内在信号和条件信号来流式传输校准后的置信状态；以及（ii）策略级别，利用该状态自适应地终止低置信度轨迹，并动态调度指导以实现阶段适宜的条件强度。在通用和组合基准测试上的实验表明，ScalingAR（1）在GenEval上将基础模型性能提升了12.5%，在TIIF-Bench上提升了15.2%；（2）在超越基线模型的同时，高效地将视觉令牌消耗降低了62.0%；以及（3）成功增强了鲁棒性，在具有挑战性的场景中将性能下降缓解了26.0%。