arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2510.03597cs.GRcs.AIcs.LG
收藏

Neon:来自自训练的负外推提升图像生成

Neon: Negative Extrapolation From Self-Training Improves Image Generation

The University of Texas at Austin(德克萨斯大学奥斯汀分校);Rice University(莱斯大学)

Sina Alemohammad, Zhangyang Wang, Richard G. Baraniuk

AI总结:

针对生成模型自训练导致模型崩溃的问题,提出Neon方法,通过反转梯度外推纠正数据分布偏差,仅需少量合成样本和计算即可提升图像生成质量,在ImageNet上达到SOTA FID 1.02。

摘要:

扩展生成式AI模型受限于高质量训练数据的稀缺性。从生成模型合成数据的便捷性表明,可以使用(未经验证的)合成数据来扩充有限真实数据语料库,以进行微调,期望提升性能。然而,不幸的是,由此产生的正反馈循环会导致模型自噬障碍(MAD,又称模型崩溃),表现为样本质量和/或多样性的快速退化。在本文中,我们提出Neon(Negative Extrapolation frOm self-traiNing),一种新的学习方法,将自训练中的退化转化为自我改进的强大信号。给定一个基础模型,Neon首先在其自合成数据上对其进行微调,但随后,反直觉地,反转其梯度更新,以从退化权重外推。我们证明Neon有效,因为典型的推理采样器偏好高概率区域,在合成数据和真实数据群体梯度之间产生可预测的反对齐,负外推纠正了这一点,使模型更好地与真实数据分布对齐。Neon实现非常简单,通过简单的后处理合并,不需要新的真实数据,仅需少至1k个合成样本即可有效工作,且通常使用不到1%的额外训练计算量。我们展示了Neon在多种架构(扩散模型、流匹配、自回归和归纳矩匹配模型)和数据集(ImageNet、CIFAR-10和FFHQ)上的普适性。特别是,在ImageNet 256x256上,Neon将xAR-L模型提升到新的最先进FID 1.02,仅增加0.36%的训练计算量。代码可在https://github.com/VITA-Group/Neon获取。

英文摘要:

Scaling generative AI models is bottlenecked by the scarcity of high-quality training data. The ease of synthesizing from a generative model suggests using (unverified) synthetic data to augment a limited corpus of real data for the purpose of fine-tuning in the hope of improving performance. Unfortunately, however, the resulting positive feedback loop leads to model autophagy disorder (MAD, aka model collapse) that results in a rapid degradation in sample quality and/or diversity. In this paper, we introduce Neon (for Negative Extrapolation frOm self-traiNing), a new learning method that turns the degradation from self-training into a powerful signal for self-improvement. Given a base model, Neon first fine-tunes it on its own self-synthesized data but then, counterintuitively, reverses its gradient updates to extrapolate away from the degraded weights. We prove that Neon works because typical inference samplers that favor high-probability regions create a predictable anti-alignment between the synthetic and real data population gradients, which negative extrapolation corrects to better align the model with the true data distribution. Neon is remarkably easy to implement via a simple post-hoc merge that requires no new real data, works effectively with as few as 1k synthetic samples, and typically uses less than 1% additional training compute. We demonstrate Neon's universality across a range of architectures (diffusion, flow matching, autoregressive, and inductive moment matching models) and datasets (ImageNet, CIFAR-10, and FFHQ). In particular, on ImageNet 256x256, Neon elevates the xAR-L model to a new state-of-the-art FID of 1.02 with only 0.36% additional training compute. Code is available at https://github.com/VITA-Group/Neon

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。

↑