Moral Stories: Situated Reasoning about Norms, Intents, Actions, and their Consequences
Denis Emelin, Ronan Le Bras, Jena D. Hwang, Maxwell Forbes, Yejin Choi
TL;DR
This work introduces Moral Stories, a crowdsourced dataset of grounded, goal-directed moral narratives with explicit normative constraints. It evaluates both classification and generation with varying degrees of grounding and proposes Chain-of-Experts decoding to enforce normative alignment. The results show that rich grounding improves task performance and that abductive reasoning and synthetic consequence signals bolster norm discovery and moral action generation. Overall, the study demonstrates the potential of structured narratives and expert-augmented decoding to steer AI behavior in socially appropriate directions, while outlining avenues for broader norm coverage and ethical safeguards.
Abstract
In social settings, much of human behavior is governed by unspoken rules of conduct. For artificial systems to be fully integrated into social environments, adherence to such norms is a central prerequisite. We investigate whether contemporary NLG models can function as behavioral priors for systems deployed in social settings by generating action hypotheses that achieve predefined goals under moral constraints. Moreover, we examine if models can anticipate likely consequences of (im)moral actions, or explain why certain actions are preferable by generating relevant norms. For this purpose, we introduce 'Moral Stories', a crowd-sourced dataset of structured, branching narratives for the study of grounded, goal-oriented social reasoning. Finally, we propose decoding strategies that effectively combine multiple expert models to significantly improve the quality of generated actions, consequences, and norms compared to strong baselines, e.g. though abductive reasoning.