Language Models Represent Beliefs of Self and Others

Autor:	Zhu, Wentao, Zhang, Zhining, Wang, Yizhou
Rok vydání:	2024
Předmět:	Computer Science - Artificial Intelligence Computer Science - Computation and Language
Druh dokumentu:	Working Paper
Popis:	Understanding and attributing mental states, known as Theory of Mind (ToM), emerges as a fundamental capability for human social reasoning. While Large Language Models (LLMs) appear to possess certain ToM abilities, the mechanisms underlying these capabilities remain elusive. In this study, we discover that it is possible to linearly decode the belief status from the perspectives of various agents through neural activations of language models, indicating the existence of internal representations of self and others' beliefs. By manipulating these representations, we observe dramatic changes in the models' ToM performance, underscoring their pivotal role in the social reasoning process. Additionally, our findings extend to diverse social reasoning tasks that involve different causal inference patterns, suggesting the potential generalizability of these representations. Comment: project page: https://walter0807.github.io/RepBelief/
Databáze:	arXiv
Externí odkaz:	http://arxiv.org/abs/2402.18496 Zobrazit plný text záznamu View this record from Arxiv