Computer Science > Computation and Language
[Submitted on 7 Sep 2021 (this version), latest version 8 Jun 2022 (v3)]
Title:Rare Words Degenerate All Words
View PDFAbstract:Despite advances in neural network language model, the representation degeneration problem of embeddings is still challenging. Recent studies have found that the learned output embeddings are degenerated into a narrow-cone distribution which makes the similarity between each embeddings positive. They analyzed the cause of the degeneration problem has been demonstrated as common to most embeddings. However, we found that the degeneration problem is especially originated from the training of embeddings of rare words. In this study, we analyze the intrinsic mechanism of the degeneration of rare word embeddings with respect of their gradient about the negative log-likelihood loss function. Furthermore, we theoretically and empirically demonstrate that the degeneration of rare word embeddings causes the degeneration of non-rare word embeddings, and that the overall degeneration problem can be alleviated by preventing the degeneration of rare word embeddings. Based on our analyses, we propose a novel method, Adaptive Gradient Partial Scaling(AGPS), to address the degeneration problem. Experimental results demonstrate the effectiveness of the proposed method qualitatively and quantitatively.
Submission history
From: Sangwon Yu [view email][v1] Tue, 7 Sep 2021 14:48:12 UTC (1,591 KB)
[v2] Wed, 16 Mar 2022 09:24:20 UTC (1,109 KB)
[v3] Wed, 8 Jun 2022 08:37:39 UTC (1,109 KB)
References & Citations
Bibliographic and Citation Tools
Bibliographic Explorer (What is the Explorer?)
Connected Papers (What is Connected Papers?)
Litmaps (What is Litmaps?)
scite Smart Citations (What are Smart Citations?)
Code, Data and Media Associated with this Article
alphaXiv (What is alphaXiv?)
CatalyzeX Code Finder for Papers (What is CatalyzeX?)
DagsHub (What is DagsHub?)
Gotit.pub (What is GotitPub?)
Hugging Face (What is Huggingface?)
Papers with Code (What is Papers with Code?)
ScienceCast (What is ScienceCast?)
Demos
Recommenders and Search Tools
Influence Flower (What are Influence Flowers?)
CORE Recommender (What is CORE?)
arXivLabs: experimental projects with community collaborators
arXivLabs is a framework that allows collaborators to develop and share new arXiv features directly on our website.
Both individuals and organizations that work with arXivLabs have embraced and accepted our values of openness, community, excellence, and user data privacy. arXiv is committed to these values and only works with partners that adhere to them.
Have an idea for a project that will add value for arXiv's community? Learn more about arXivLabs.