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 Tools annual

Global: Zendy Free Plan

Global: Zendy Tools monthly

Global: Zendy Plus monthly

Large Language Models have shown exceptional generative abilities in variousnatural language and generation tasks. However, possible anthropomorphizationand leniency towards failure cases have propelled discussions on emergentabilities of Large Language Models especially on Theory of Mind (ToM) abilitiesin Large Language Models. While several false-belief tests exists to verify theability to infer and maintain mental models of another entity, we study aspecial application of ToM abilities that has higher stakes and possiblyirreversible consequences : Human Robot Interaction. In this work, we explorethe task of Perceived Behavior Recognition, where a robot employs a LargeLanguage Model (LLM) to assess the robot's generated behavior in a mannersimilar to human observer. We focus on four behavior types, namely -explicable, legible, predictable, and obfuscatory behavior which have beenextensively used to synthesize interpretable robot behaviors. The LLMs goal is,therefore to be a human proxy to the agent, and to answer how a certain agentbehavior would be perceived by the human in the loop, for example "Given arobot's behavior X, would the human observer find it explicable?". We conduct ahuman subject study to verify that the users are able to correctly answer sucha question in the curated situations (robot setting and plan) across fivedomains. A first analysis of the belief test yields extremely positive resultsinflating ones expectations of LLMs possessing ToM abilities. We then proposeand perform a suite of perturbation tests which breaks this illusion, i.e.Inconsistent Belief, Uninformative Context and Conviction Test. We concludethat, the high score of LLMs on vanilla prompts showcases its potential use inHRI settings, however to possess ToM demands invariance to trivial orirrelevant perturbations in the context which LLMs lack.

Theory of Mind abilities of Large Language Models in Human-Robot Interaction : An Illusion?