Distributed Partial Information Puzzles: Examining Common Ground Construction Under Epistemic Asymmetry
This work addresses the challenge of establishing common ground for AI systems in multimodal, multiparty settings with asymmetric information, which is a foundational problem for collaborative AI.
This paper introduces the Distributed Partial Information Puzzle (DPIP), a collaborative construction task designed to study common ground establishment under epistemic asymmetry in AI systems. They created a multimodal dataset from these interactions and found that state-of-the-art large language models struggle to track task progression and belief states compared to an axiomatic pipeline grounded in Dynamic Epistemic Logic.
Establishing common ground, a shared set of beliefs and mutually recognized facts, is fundamental to collaboration, yet remains a challenge for current AI systems, especially in multimodal, multiparty settings, where the collaborators bring different information to the table. We introduce the Distributed Partial Information Puzzle (DPIP), a collaborative construction task that elicits rich multimodal communication under epistemic asymmetry. We present a multimodal dataset of these interactions, annotated and temporally aligned across speech, gesture, and action modalities to support reasoning over propositional content and belief dynamics. We then evaluate two paradigms for modeling common ground (CG): (1) state-of-the-art large language models (LLMs), prompted to infer shared beliefs from multimodal updates, and (2) an axiomatic pipeline grounded in Dynamic Epistemic Logic (DEL) that incrementally performs the same task. Results on the annotated DPIP data indicate that it poses a challenge to modern LLMs' abilities to track both task progression and belief state.