InfoCL: Alleviating Catastrophic Forgetting in Continual Text Classification from An Information Theoretic Perspective
Open Access
- 1 January 2023
- conference paper
- Published by Association for Computational Linguistics (ACL)
- p. 14557-14570
Abstract
Yifan Song, Peiyi Wang, Weimin Xiong, Dawei Zhu, Tianyu Liu, Zhifang Sui, Sujian Li. Findings of the Association for Computational Linguistics: EMNLP 2023. 2023.Keywords
All Related Versions
This publication has 25 references indexed in Scilit:
- Co2L: Contrastive Continual Learning2021
- Benchmarking Natural Language Understanding Services for Building Conversational Agents2021
- A continual learning survey: Defying forgetting in classification tasksIEEE Transactions on Pattern Analysis and Machine Intelligence, 2021
- GDumb: A Simple Approach that Questions Our Progress in Continual Learning2020
- Momentum Contrast for Unsupervised Visual Representation Learning2020
- Memory Aware Synapses: Learning What (not) to Forget2018
- PackNet: Adding Multiple Tasks to a Single Network by Iterative Pruning2018
- Emergence of Invariance and Disentanglement in Deep Representations2018
- iCaRL: Incremental Classifier and Representation Learning2017
- Deep learning and the information bottleneck principle2015