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

Learned communication between agents is a powerful tool when approachingdecision-making problems that are hard to overcome by any single agent inisolation. However, continual coordination and communication learning betweenmachine agents or human-machine partnerships remains a challenging openproblem. As a stepping stone toward solving the continual communicationlearning problem, in this paper we contribute a multi-faceted study into whatwe term Pavlovian signalling -- a process by which learned, temporally extendedpredictions made by one agent inform decision-making by another agent withdifferent perceptual access to their shared environment. We seek to establishhow different temporal processes and representational choices impact Pavloviansignalling between learning agents. To do so, we introduce a partiallyobservable decision-making domain we call the Frost Hollow. In this domain aprediction learning agent and a reinforcement learning agent are coupled into atwo-part decision-making system that seeks to acquire sparse reward whileavoiding time-conditional hazards. We evaluate two domain variations: 1)machine prediction and control learning in a linear walk, and 2) a predictionlearning machine interacting with a human participant in a virtual realityenvironment. Our results showcase the speed of learning for Pavloviansignalling, the impact that different temporal representations do (and do not)have on agent-agent coordination, and how temporal aliasing impacts agent-agentand human-agent interactions differently. As a main contribution, we establishPavlovian signalling as a natural bridge between fixed signalling paradigms andfully adaptive communication learning. Our results therefore point to anactionable, constructivist path towards continual communication learningbetween reinforcement learning agents, with potential impact in a range ofreal-world settings.

The Frost Hollow Experiments: Pavlovian Signalling as a Path to Coordination and Communication Between Agents