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

Value decomposition is widely used in cooperative multi-agent reinforcementlearning, however, its implicit credit assignment mechanism is not yet fullyunderstood due to black-box networks. In this work, we study an interpretablevalue decomposition framework via the family of generalized additive models. Wepresent a novel method, named Neural Attention Additive Q-learning(N$\text{A}^\text{2}$Q), providing inherent intelligibility of collaborationbehavior. N$\text{A}^\text{2}$Q can explicitly factorize the optimal jointpolicy induced by enriching shape functions to model all possible coalitions ofagents into individual policies. Moreover, we construct identity semantics topromote estimating credits together with the global state and individual valuefunctions, where local semantic masks help us diagnose whether each agentcaptures relevant-task information. Extensive experiments show thatN$\text{A}^\text{2}$Q consistently achieves superior performance compared todifferent state-of-the-art methods on all challenging tasks, while yieldinghuman-like interpretability.

N$\text{A}^\text{2}$Q: Neural Attention Additive Model for Interpretable Multi-Agent Q-Learning