Explaining and Improving Information Complementarities in Multi-Agent Decision-making

Ziyang Guo, Yifan Wu, Jason Hartline, Jessica Hullman · 2025

Read Explaining and Improving Information Complementarities in Multi-Agent Decision-making by Ziyang Guo, Yifan Wu, Jason Hartline, Jessica Hullman (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-2502-06152.txt.

Description

Ziyang Guo and co-authors build a way to show what each partner in a human-AI team uniquely knows, in this 2025 paper. Teams are picked in hopes their strengths combine, but improving them means knowing which facts each side holds and uses. The authors frame the value of information with decision theory, aimed above all at people working with AI.

Their explanation tool adapts a popular method to spotlight facts that complement the human partner rather than repeat the AI's own logic. In a study of human-AI decisions the tool's displays cut errors more reliably than the standard displays when paired with AI predictions. The team also demos the framework on chest X-ray reading and deepfake spotting. The pitch is practical: show people the missing piece, not the whole puzzle again.

Catalog record

CREATORZiyang Guo, Yifan Wu, Jason Hartline, Jessica Hullman
YEAR2025
COLLECTIONPAPERS
FORMATTXT
FILE SIZE76 KB
VERSION1.0
LICENSEhttp://creativecommons.org/licenses/by/4.0/
RIGHTSCC BY
SOURCEarXiv

Source: https://arxiv.org/abs/2502-06152

Browse the full library · About the Tower of Babel