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

Lifelong learning is a very important step toward realizing robust autonomousartificial agents. Neural networks are the main engine of deep learning, whichis the current state-of-the-art technique in formulating adaptive artificialintelligent systems. However, neural networks suffer from catastrophicforgetting when stressed with the challenge of continual learning. Weinvestigate how to exploit modular topology in neural networks in order todynamically balance the information load between different modules by routinginputs based on the information content in each module so that informationinterference is minimized. Our dynamic information balancing (DIB) techniqueadapts a reinforcement learning technique to guide the routing of differentinputs based on a reward signal derived from a measure of the information loadin each module. Our empirical results show that DIB combined with elasticweight consolidation (EWC) regularization outperforms models with similarcapacity and EWC regularization across different task formulations anddatasets.

Reducing Catastrophic Forgetting in Modular Neural Networks by Dynamic Information Balancing