English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

Continual learning (CL) aims to constantly learn new knowledge over timewhile avoiding catastrophic forgetting on old tasks. We focus on continual textclassification under the class-incremental setting. Recent CL studies haveidentified the severe performance decrease on analogous classes as a key factorfor catastrophic forgetting. In this paper, through an in-depth exploration ofthe representation learning process in CL, we discover that the compressioneffect of the information bottleneck leads to confusion on analogous classes.To enable the model learn more sufficient representations, we propose a novelreplay-based continual text classification method, InfoCL. Our approachutilizes fast-slow and current-past contrastive learning to perform mutualinformation maximization and better recover the previously learnedrepresentations. In addition, InfoCL incorporates an adversarial memoryaugmentation strategy to alleviate the overfitting problem of replay.Experimental results demonstrate that InfoCL effectively mitigates forgettingand achieves state-of-the-art performance on three text classification tasks.The code is publicly available at https://github.com/Yifan-Song793/InfoCL.

InfoCL: Alleviating Catastrophic Forgetting in Continual Text Classification from An Information Theoretic Perspective