BEDA: Belief Estimation as Probabilistic Constraints for Performing Strategic Dialogue Acts

TL;DR

This paper introduces BEDA, a framework that uses probabilistic belief constraints to improve strategic dialogue generation across various settings, leading to significant performance gains.

Contribution

It formalizes belief-based probabilistic constraints for dialogue acts and demonstrates their effectiveness in enhancing strategic dialogue performance.

Findings

BEDA outperforms baselines on CKBG, improving success rate by up to 20.6 points.

BEDA achieves an average of 9.3 points improvement on Mutual Friends.

On CaSiNo, BEDA attains the best deal compared to all baselines.

Abstract

Strategic dialogue requires agents to execute distinct dialogue acts, for which belief estimation is essential. While prior work often estimates beliefs accurately, it lacks a principled mechanism to use those beliefs during generation. We bridge this gap by first formalizing two core acts Adversarial and Alignment, and by operationalizing them via probabilistic constraints on what an agent may generate. We instantiate this idea in BEDA, a framework that consists of the world set, the belief estimator for belief estimation, and the conditional generator that selects acts and realizes utterances consistent with the inferred beliefs. Across three settings, Conditional Keeper Burglar (CKBG, adversarial), Mutual Friends (MF, cooperative), and CaSiNo (negotiation), BEDA consistently outperforms strong baselines: on CKBG it improves success rate by at least 5.0 points across backbones and by 20.6 points with GPT-4.1-nano; on Mutual Friends it achieves an average improvement of 9.3 points; and on CaSiNo it achieves the optimal deal relative to all baselines. These results indicate that casting belief estimation as constraints provides a simple, general mechanism for reliable strategic dialogue.