AsymPuzl: An Asymmetric Puzzle for multi-agent cooperation

Xavier Cadet, Edward Koh, Peter Chin · 2025

Read AsymPuzl: An Asymmetric Puzzle for multi-agent cooperation by Xavier Cadet, Edward Koh, Peter Chin (2025) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2512-03466.txt.

Description

Two agents with half the facts each must cooperate to win. Xavier Cadet, Edward Koh, and Peter Chin present AsymPuzl in this 2025 work. Most multi-agent setups use open-ended role-play, which is hard to evaluate in a controlled way.

AsymPuzl is a minimal two-agent puzzle with information asymmetry. Each agent sees a partial, complementary view of a symbolic puzzle. They must exchange messages to solve it together. The setup isolates communication skill, a clean target for testing language model teamwork. Minimal puzzles strip away role-play noise, leaving pure communication skill exposed.

Catalog record

CREATORXavier Cadet, Edward Koh, Peter Chin
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE25 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2512-03466

Browse the full library · About the Tower of Babel