analiza aiqbench
This research addresses a critical gap in large language models by using pragmatic inference to better align stated text with underlying moral implications. By leveraging metapragmatic links, the study demonstrates improved generalization across diverse tasks without relying on external datasets for every scenario. This work advances AI safety and alignment efforts by enabling systems to more robustly navigate complex social variables inherent in ethical decision-making.
Pragmatic Inference for Moral Reasoning Acquisition: Generalization via Metapragmatic Links
arXiv:2509.24102v5 Announce Type: replace Abstract: While moral reasoning has emerged as a promising research direction for large language models (LLMs), achieving robust generalization remains a critical challenge. This challenge arises from the gap between what is said and what is morally implied. In this paper, we build on metapragmatic links and Moral Foundations Theory to close this gap. Specifically, we develop a pragmatic inference approach that enables LLMs, given a moral situation, to acquire the metapragmatic links between moral reasoning objectives and the social variables that influence them. We adapt this approach to three different moral reasoning tasks to demonstrate its adaptability and generalizability. Experimental results show that our approach significantly enhances LLMs' generalization in moral reasoning, paving the way for future research to leverage pragmatic inference across a wide range of moral reasoning tasks.
Abstract:While moral reasoning has emerged as a promising research direction for large language models (LLMs), achieving robust generalization remains a critical challenge. This challenge arises from the gap between what is said and what is morally implied. In this paper, we build on metapragmatic links and Moral Foundations Theory to close this gap. Specifically, we develop a pragmatic inference approach that enables LLMs, given a moral situation, to acquire the metapragmatic links between moral reasoning objectives and the social variables that influence them. We adapt this approach to three different moral reasoning tasks to demonstrate its adaptability and generalizability. Experimental results show that our approach significantly enhances LLMs' generalization in moral reasoning, paving the way for future research to leverage pragmatic inference across a wide range of moral reasoning tasks.
Computation and Language (cs.CL)
arXiv:2509.24102 [cs.CL]
(or arXiv:2509.24102v5 [cs.CL] for this version)
Text integral sub licența deschisă a publicației · 142 cuvinte