arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2610.36962cs.AI
收藏

机器人会感到不适吗?评估大型语言模型(LLMs)中功能效价的经验证据

Do bots feel bad? Assessing empirical evidence for functional valence in LLMs

University of Copenhagen(哥本哈根大学)

Iwan Williams

AI总结:

本文评估LLMs功能效价的现有经验证据,指出关键效价方面未被测试、现有范式无法区分效价与邻近构念,并提出新实验方法以精准研究功能效价。

摘要:

大型语言模型(LLMs)是否具有效价状态,即感到愉悦或不适的状态?近期的经验研究声称表明LLMs拥有功能效价状态,例如“功能愉悦与痛苦”(Ren等人,2026)、“功能情绪”(Sofroniew等人,2026)以及“功能福利”(Han等人,2026),这些状态在认知与行动中发挥着与效价状态相同的因果作用,且不一定需要有意识的体验。但目前尚未探究这些已识别的状态在多大程度上符合哲学与认知科学文献中所理解的效价功能角色。本文评估现有证据,提出两点:其一,效价的关键方面在LLMs中仍未得到测试;其二,现有经验范式无法充分区分(功能)效价与偏好、价值判断、效价归因及角色扮演等邻近构念。最后,本文勾勒出一些新的实验方法,以更好地针对(功能)效价本身进行研究。

英文摘要:

Do LLMs have valenced states, i.e. states that feel pleasant or unpleasant? Recent empirical work has claimed to show that LLMs possess functional valenced states (e.g. ``functional pleasure and pain'' [Ren et al. 2026], ``functional emotions'' [Sofroniew et al. 2026], and ``functional welfare'' [Han et al. 2026]) that play the same causal role as valenced states in cognition and action, without necessarily being consciously experienced. But it remains unexplored how well the identified states fit the functional role of valence as understood in the philosophical and cognitive science literature. Here, I assess the current body of evidence, and argue (i) that key aspects of valence remain untested in LLMs, and (ii) that extant empirical paradigms inadequately distinguish (functional) valence from neighbouring constructs like preferences, value-judgments, valence attributions and role-play. I conclude by sketching some novel experimental approaches to better target (functional) valence itself.

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。

↑