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

Unlike perfect information games, where all elements are known to everyplayer, imperfect information games emulate the real-world complexities ofdecision-making under uncertain or incomplete information. GPT-4, the recentbreakthrough in large language models (LLMs) trained on massive passive data,is notable for its knowledge retrieval and reasoning abilities. This paperdelves into the applicability of GPT-4's learned knowledge for imperfectinformation games. To achieve this, we introduce \textbf{Suspicion-Agent}, aninnovative agent that leverages GPT-4's capabilities for performing inimperfect information games. With proper prompt engineering to achievedifferent functions, Suspicion-Agent based on GPT-4 demonstrates remarkableadaptability across a range of imperfect information card games. Importantly,GPT-4 displays a strong high-order theory of mind (ToM) capacity, meaning itcan understand others and intentionally impact others' behavior. Leveragingthis, we design a planning strategy that enables GPT-4 to competently playagainst different opponents, adapting its gameplay style as needed, whilerequiring only the game rules and descriptions of observations as input. In theexperiments, we qualitatively showcase the capabilities of Suspicion-Agentacross three different imperfect information games and then quantitativelyevaluate it in Leduc Hold'em. The results show that Suspicion-Agent canpotentially outperform traditional algorithms designed for imperfectinformation games, without any specialized training or examples. In order toencourage and foster deeper insights within the community, we make ourgame-related data publicly available.

Suspicion-Agent: Playing Imperfect Information Games with Theory of Mind Aware GPT-4