arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2202.01334cs.LGcs.AI
收藏

基于动态矢量Quantization的自适应离散通信瓶颈

Adaptive Discrete Communication Bottlenecks with Dynamic Vector Quantization

Mila(米拉);Google(谷歌);National University of Singapore(新加坡国立大学)

Dianbo Liu, Alex Lamb, Xu Ji, Pascal Notsawo, Mike Mozer, Yoshua Bengio, Kenji Kawaguchi

摘要:

矢量量化(VQ)是一种离散化潜在表示的方法,已成为深度学习工具包的主要组成部分。理论和经验表明,表示的离散化导致改进的泛化,包括在强化学习中,其中离散化可以用于瓶颈多智能体通信,以促进代理专业化和鲁棒性。大多数基于矢量量化的方法的离散化紧密度由表示矢量中的离散码的数量和码本大小定义,它们被固定为超参数。在这项工作中,我们建议学习动态选择离散化的输入条件下的紧密性,基于假设数据自然包含的复杂性,要求不同程度的代表性粗糙度的变化。我们表明,动态变化的紧密性的通信瓶颈可以提高模型的视觉推理和强化学习任务的性能。

英文摘要:

Vector Quantization (VQ) is a method for discretizing latent representations and has become a major part of the deep learning toolkit. It has been theoretically and empirically shown that discretization of representations leads to improved generalization, including in reinforcement learning where discretization can be used to bottleneck multi-agent communication to promote agent specialization and robustness. The discretization tightness of most VQ-based methods is defined by the number of discrete codes in the representation vector and the codebook size, which are fixed as hyperparameters. In this work, we propose learning to dynamically select discretization tightness conditioned on inputs, based on the hypothesis that data naturally contains variations in complexity that call for different levels of representational coarseness. We show that dynamically varying tightness in communication bottlenecks can improve model performance on visual reasoning and reinforcement learning tasks.

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。

↑