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

Over-generalization is a thorny issue in cognitive science, where people maybecome overly cautious due to past experiences. Agents in multi-agentreinforcement learning (MARL) also have been found to suffer relativeover-generalization (RO) as people do and stuck to sub-optimal cooperation.Recent methods have shown that assigning reasoning ability to agents canmitigate RO algorithmically and empirically, but there has been a lack oftheoretical understanding of RO, let alone designing provably RO-free methods.This paper first proves that RO can be avoided when the MARL method satisfies aconsistent reasoning requirement under certain conditions. Then we introduce anovel reasoning framework, called negotiated reasoning, that first builds theconnection between reasoning and RO with theoretical justifications. Afterthat, we propose an instantiated algorithm, Stein variational negotiatedreasoning (SVNR), which uses Stein variational gradient descent to derive anegotiation policy that provably avoids RO in MARL under maximum entropy policyiteration. The method is further parameterized with neural networks foramortized learning, making computation efficient. Numerical experiments on manyRO-challenged environments demonstrate the superiority and efficiency of SVNRcompared to state-of-the-art methods in addressing RO.

Negotiated Reasoning: On Provably Addressing Relative Over-Generalization