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

Human culture relies on innovation: our ability to continuously explore howexisting elements can be combined to create new ones. Innovation is notsolitary, it relies on collective search and accumulation. Reinforcementlearning (RL) approaches commonly assume that fully-connected groups are bestsuited for innovation. However, human laboratory and field studies have shownthat hierarchical innovation is more robustly achieved by dynamic socialnetwork structures. In dynamic settings, humans oscillate between innovatingindividually or in small clusters, and then sharing outcomes with others. Toour knowledge, the role of social network structure on innovation has not beensystematically studied in RL. Here, we use a multi-level problem setting(WordCraft), with three different innovation tasks to test the hypothesis thatthe social network structure affects the performance of distributed RLalgorithms. We systematically design networks of DQNs sharing experiences fromtheir replay buffers in varying structures (fully-connected, small world,dynamic, ring) and introduce a set of behavioral and mnemonic metrics thatextend the classical reward-focused evaluation framework of RL. Comparing thelevel of innovation achieved by different social network structures acrossdifferent tasks shows that, first, consistent with human findings, experiencesharing within a dynamic structure achieves the highest level of innovation intasks with a deceptive nature and large search spaces. Second, experiencesharing is not as helpful when there is a single clear path to innovation.Third, the metrics we propose, can help understand the success of differentsocial network structures on different tasks, with the diversity of experienceson an individual and group level lending crucial insights.

Social Network Structure Shapes Innovation: Experience-sharing in RL with SAPIENS