# Rationality-Robust Information Design: Bayesian Persuasion under Quantal Response

arXiv: https://arxiv.org/abs/2207.08253 · HTML: https://mlsanity.com/paper/2207.08253/

## Key facts

- Cited by 6 papers in the map (data through 14 Oct 2026).
- First posted on arXiv 17 Jul 2022.

## Built on by (most cited first, 6 of 6)

- [Persuading a Behavioral Agent: Approximately Best Responding and Learning](https://mlsanity.com/paper/2302.03719/) (2023)
- [Generalized Principal-Agent Problem with a Learning Agent](https://mlsanity.com/paper/2402.09721/) (2024)
- [Goal Recognition Design for General Behavioral Agents using Machine Learning](https://mlsanity.com/paper/2404.03054/) (2024)
- [Make an Offer They Can't Refuse: Grounding Bayesian Persuasion in Real-World Dialogues without Pre-Commitment](https://mlsanity.com/paper/2510.13387/) (2025)
- [On the Utility of Accounting for Human Beliefs about AI Intention in Human-AI Collaboration](https://mlsanity.com/paper/2406.06051/) (2024)
- [One AI Signal, Many Human Judgments: A Bayesian Cascade Analysis of AI-based Credibility Indicators in Online Information Spread](https://mlsanity.com/paper/2608.30311/) (2026)

## Abstract (arXiv)

Classic mechanism/information design imposes the assumption that agents are fully rational, meaning each of them always selects the action that maximizes her expected utility. Yet many empirical evidence suggests that human decisions may deviate from this full rationality assumption. In this work, we attempt to relax the full rationality assumption with bounded rationality. Specifically, we formulate the bounded rationality of an agent by adopting the quantal response model (McKelvey and Palfrey, 1995). We develop a theory of rationality-robust information design in the canonical setting of Bayesian persuasion (Kamenica and Gentzkow, 2011) with binary receiver action. We first identify conditions under which the optimal signaling scheme structure for a fully rational receiver remains optimal or approximately optimal for a boundedly rational receiver. In practice, it might be costly for the designer to estimate the degree of the receiver's bounded rationality level. Motivated by this practical consideration, we then study the existence and construction of robust signaling schemes when there is uncertainty about the receiver's bounded rationality level.