arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2104.08197cs.LGcs.CL
收藏

探索人工神经网络:来自神经科学的见解

Probing artificial neural networks: insights from neuroscience

MIT(麻省理工学院);Stanford University(斯坦福大学)

Anna A. Ivanova, John Hewitt, Noga Zaslavsky

摘要:

神经科学和机器学习的一个主要挑战是开发有用的工具来理解复杂的信息处理系统。一种这样的工具是探针,即,将感兴趣的特征与生物或人工神经网络中出现的激活模式相关联的监督模型。神经科学通过近几十年来进行的大量研究为使用这些模型铺平了道路。在这项工作中,我们从神经科学中汲取见解,以帮助指导机器学习的探索研究。我们强调两个重要的设计选择探针$-$的方向和表现力$-$,并与这些选择的研究目标。我们认为,具体的研究目标发挥了至关重要的作用时,设计一个探针,并鼓励未来的探测研究明确说明这些目标。

英文摘要:

A major challenge in both neuroscience and machine learning is the development of useful tools for understanding complex information processing systems. One such tool is probes, i.e., supervised models that relate features of interest to activation patterns arising in biological or artificial neural networks. Neuroscience has paved the way in using such models through numerous studies conducted in recent decades. In this work, we draw insights from neuroscience to help guide probing research in machine learning. We highlight two important design choices for probes $-$ direction and expressivity $-$ and relate these choices to research goals. We argue that specific research goals play a paramount role when designing a probe and encourage future probing studies to be explicit in stating these goals.

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。

↑