arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2602.02467cs.CL
收藏

信念引导的代理与元认知监控在大语言模型中的迹象

Indications of Belief-Guided Agency and Meta-Cognitive Monitoring in Large Language Models

Blavatnik School of Computer Science and AI(Blavatnik计算机科学与人工智能学院);Tel Aviv University(特拉维夫大学);School of Psychological Sciences(心理学学院);Sagol School of Neuroscience(神经科学学院);Canadian Institute for Advanced Research (CIFAR)(加拿大高级研究 institute);University of Cambridge(剑桥大学)

Noam Steinmetz Yalon, Ariel Goldstein, Liad Mudrik, Mor Geva

AI总结:

本研究通过HOT-3指标探讨大语言模型中信念引导的代理与元认知监控的存在,揭示了信念形成对行动选择的影响及模型自我监控的能力。

摘要:

大语言模型(LLMs)的快速进步引发了关于这些模型是否具备某种意识形式的问题。为解决这一挑战,Butlin等人(2023)基于神经科学理论提出了一套意识在人工系统中的指标。在本工作中,我们评估了该列表中的一个关键指标,称为HOT-3,该指标测试了由一般信念形成和行动选择系统引导的代理能力,该系统根据元认知监控更新信念。我们将信念视为模型潜在空间中的表示,这些表示在给定输入下产生,并引入了量化其在生成过程中主导程度的指标。分析模型和任务中竞争性信念的动力学揭示了三个关键发现:(1)外部操纵系统地调节内部信念形成;(2)信念形成因果地驱动模型的行动选择;(3)模型能够监控并报告自身的信念状态。这些结果为LLMs中信念引导的代理和元认知监控的存在提供了实证支持。更广泛地说,我们的工作为研究LLMs中代理、信念和元认知的出现奠定了方法论基础。

英文摘要:

Rapid advancements in large language models (LLMs) have sparked the question whether these models possess some form of consciousness. To tackle this challenge, Butlin et al. (2023) introduced a list of indicators for consciousness in artificial systems based on neuroscientific theories. In this work, we evaluate a key indicator from this list, called HOT-3, which tests for agency guided by a general belief-formation and action selection system that updates beliefs based on meta-cognitive monitoring. We view beliefs as representations in the model's latent space that emerge in response to a given input, and introduce a metric to quantify their dominance during generation. Analyzing the dynamics between competing beliefs across models and tasks reveals three key findings: (1) external manipulations systematically modulate internal belief formation, (2) belief formation causally drives the model's action selection, and (3) models can monitor and report their own belief states. Together, these results provide empirical support for the existence of belief-guided agency and meta-cognitive monitoring in LLMs. More broadly, our work lays methodological groundwork for investigating the emergence of agency, beliefs, and meta-cognition in LLMs.

机构、总结、中文摘要均由AI分析而来,可能存在错误,请以论文正文为准。

↑