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

There is a recent trend of applying multi-agent reinforcement learning (MARL)to train an agent that can cooperate with humans in a zero-shot fashion withoutusing any human data. The typical workflow is to first repeatedly run self-play(SP) to build a policy pool and then train the final adaptive policy againstthis pool. A crucial limitation of this framework is that every policy in thepool is optimized w.r.t. the environment reward function, which implicitlyassumes that the testing partners of the adaptive policy will be preciselyoptimizing the same reward function as well. However, human objectives areoften substantially biased according to their own preferences, which can differgreatly from the environment reward. We propose a more general framework,Hidden-Utility Self-Play (HSP), which explicitly models human biases as hiddenreward functions in the self-play objective. By approximating the reward spaceas linear functions, HSP adopts an effective technique to generate an augmentedpolicy pool with biased policies. We evaluate HSP on the Overcooked benchmark.Empirical results show that our HSP method produces higher rewards thanbaselines when cooperating with learned human models, manually scriptedpolicies, and real humans. The HSP policy is also rated as the most assistivepolicy based on human feedback.

Learning Zero-Shot Cooperation with Humans, Assuming Humans Are Biased