Verbalized Bayesian Persuasion
Wenhao Li, Yue Lin, Xiangfeng Wang, Bo Jin, Hongyuan Zha, Baoxiang Wang · 2025
Read Verbalized Bayesian Persuasion by Wenhao Li, Yue Lin, Xiangfeng Wang, Bo Jin, Hongyuan Zha, Baoxiang Wang (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-2502-01587.txt.
Description
Persuasion is a game of shared information: a sender who knows more chooses what to reveal so a receiver acts as the sender hopes. Wenhao Li and five co-authors bring language models into that classic game theory setup in this 2025 paper. Older methods coded the shared information as numbers, which confined studies to toy games far from real talk.
Their framework, verbalized Bayesian persuasion, turns the setup into a dialogue game where models play both sender and receiver. A solver mixes model talk with game math, guided by stated commitment rules, obedience checks, and deliberate hiding of some facts. Tests in settings like recommendation letters, courtrooms, and police work matched classic theory where it applies. The setup also found working persuasion plans in longer, messier exchanges.
Catalog record
| CREATOR | Wenhao Li, Yue Lin, Xiangfeng Wang, Bo Jin, Hongyuan Zha, Baoxiang Wang |
|---|---|
| YEAR | 2025 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 156 KB |
| VERSION | 1.0 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2502-01587