arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2206.12840cs.LGcs.CL
收藏

如果将自回归生成模型视为基于能量的模型,它会更好

Your Autoregressive Generative Model Can be Better If You Treat It as an Energy-Based One

Mila-Quebec AI Institute(米拉-魁北克人工智能研究所);U. Montreal(蒙特利尔大学);Nanyang Technological University(南洋理工大学);Microsoft Research Asia(微软亚洲研究院);University of Illinois, Urbana Champaign(伊利诺伊大学厄巴纳-香槟分校)

Yezhen Wang, Tong Che, Bo Li, Kaitao Song, Hengzhi Pei, Yoshua Bengio, Dongsheng Li

摘要:

自回归生成模型是常用的,特别是对于那些涉及序列数据的任务。然而,由于链式条件建模的固有特性,它们一直受到一系列固有缺陷的困扰(例如,暴露偏差或缺乏长期一致性),严重限制了它们正确建模分布的能力。在本文中,我们提出了一种称为E-ARM的独特方法,用于训练自回归生成模型,该模型利用了精心设计的基于能量的学习目标。通过利用softmax运算的额外自由度,我们可以使自回归模型本身成为一个基于能量的模型,用于测量输入的可能性,而无需引入任何额外的参数。此外,我们证明了E-ARM可以有效地训练,并且能够减轻曝光偏差问题,并增加自回归生成模型的时间相干性。广泛的实证结果,包括基准,如语言建模,神经机器翻译,图像生成,证明了所提出的方法的有效性。

英文摘要:

Autoregressive generative models are commonly used, especially for those tasks involving sequential data. They have, however, been plagued by a slew of inherent flaws due to the intrinsic characteristics of chain-style conditional modeling (e.g., exposure bias or lack of long-range coherence), severely limiting their ability to model distributions properly. In this paper, we propose a unique method termed E-ARM for training autoregressive generative models that takes advantage of a well-designed energy-based learning objective. By leveraging the extra degree of freedom of the softmax operation, we are allowed to make the autoregressive model itself be an energy-based model for measuring the likelihood of input without introducing any extra parameters. Furthermore, we show that E-ARM can be trained efficiently and is capable of alleviating the exposure bias problem and increase temporal coherence for autoregressive generative models. Extensive empirical results, covering benchmarks like language modeling, neural machine translation, and image generation, demonstrate the effectiveness of the proposed approach.

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。

↑