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 Tools annual

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 Tools monthly

Global: Zendy Plus monthly

Presents the access of premium content as premium feature

Premium Content

Global: Zendy Plus annual

Global: Zendy Free Plan

Although there are many approaches to implement intrinsically motivatedartificial agents, the combined usage of multiple intrinsic drives remainsstill a relatively unexplored research area. Specifically, we hypothesize thata mechanism capable of quantifying and controlling the evolution of theinformation flow between the agent and the environment could be the fundamentalcomponent for implementing a higher degree of autonomy into artificialintelligent agents. This paper propose a unified strategy for implementing twosemantically orthogonal intrinsic motivations: curiosity and empowerment.Curiosity reward informs the agent about the relevance of a recent agentaction, whereas empowerment is implemented as the opposite information flowfrom the agent to the environment that quantifies the agent's potential ofcontrolling its own future. We show that an additional homeostatic drive isderived from the curiosity reward, which generalizes and enhances theinformation gain of a classical curious/heterostatic reinforcement learningagent. We show how a shared internal model by curiosity and empowermentfacilitates a more efficient training of the empowerment function. Finally, wediscuss future directions for further leveraging the interplay between thesetwo intrinsic rewards.

A unified strategy for implementing curiosity and empowerment driven reinforcement learning