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 2405.08154cs.HCcs.AI
收藏

LLM 心智理论与对齐:机遇与风险

LLM Theory of Mind and Alignment: Opportunities and Risks

Google Research(谷歌研究院)

Winnie Street

AI总结:

本文围绕大语言模型(LLM)的心智理论(ToM)与人类价值观对齐的关联,从个体和群体层面梳理LLM ToM在人机交互中的应用场景及对齐的机遇风险,指明未来研究重点方向。

摘要:

大语言模型(LLM)凭借其在自然语言对话与推理方面的出色能力,正在革新人机交互模式以及人们对人工智能(AI)的认知。目前学界越来越关注LLM是否具备心智理论(ToM)——即推理他人心理与情绪状态的能力,这是人类社会智能的核心组成部分。随着LLM逐步融入个人、职业与社会生活的方方面面,并被赋予更高的决策自主权以产生具有现实影响的结果,我们亟需明确如何让其与人类价值观对齐。在这一背景下,ToM似乎是一个颇具前景的研究方向。本文参考人类ToM的作用与影响相关文献,明确了LLM ToM将在个体与群体层面的人-LLM交互中显现的关键领域,以及各领域中对齐工作面临的机遇与风险。在个体层面,本文探讨了LLM ToM可能在目标明确化、对话适配、共情与拟人化中的表现形式;在群体层面,本文分析了LLM ToM如何推动集体对齐、合作或竞争以及道德判断的形成。本文梳理了一系列潜在影响,并提出了未来研究最紧迫的方向。

英文摘要:

Large language models (LLMs) are transforming human-computer interaction and conceptions of artificial intelligence (AI) with their impressive capacities for conversing and reasoning in natural language. There is growing interest in whether LLMs have theory of mind (ToM); the ability to reason about the mental and emotional states of others that is core to human social intelligence. As LLMs are integrated into the fabric of our personal, professional and social lives and given greater agency to make decisions with real-world consequences, there is a critical need to understand how they can be aligned with human values. ToM seems to be a promising direction of inquiry in this regard. Following the literature on the role and impacts of human ToM, this paper identifies key areas in which LLM ToM will show up in human:LLM interactions at individual and group levels, and what opportunities and risks for alignment are raised in each. On the individual level, the paper considers how LLM ToM might manifest in goal specification, conversational adaptation, empathy and anthropomorphism. On the group level, it considers how LLM ToM might facilitate collective alignment, cooperation or competition, and moral judgement-making. The paper lays out a broad spectrum of potential implications and suggests the most pressing areas for future research.

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。

↑