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 2411.02432cs.CLcs.AIcs.CY
收藏

LLM能否做出涉及规定的痛苦与快乐状态的权衡?

Can LLMs make trade-offs involving stipulated pain and pleasure states?

London School of Economics(伦敦政治经济学院);Google DeepMind(谷歌DeepMind);Google(谷歌)

Geoff Keeling, Winnie Street, Martyna Stachaczyk, Daria Zakharova, Iulia M. Comsa, Anastasiya Sakovych, Isabella Logothetis, Zejia Zhang, Blaise Agüera y Arcas, Jonathan Birch

AI总结:

研究通过点数最大化游戏测试LLM在痛苦惩罚与快乐奖励下的权衡能力,发现多数模型在临界阈值后会改变决策策略,为LLM感受性辩论提供了新证据。

摘要:

快乐和痛苦在人类决策中发挥着重要作用,通过提供一种通用货币来解决动机冲突。虽然大型语言模型(LLM)能够生成关于快乐和痛苦体验的详细描述,但LLM能否在选择场景中重现快乐和痛苦的动机力量仍是一个悬而未决的问题——这一问题可能对关于LLM感受性(被理解为具有效价体验状态的能力)的辩论产生影响。我们使用一个简单的游戏来探究这个问题,该游戏声明的目标是最大化点数,但点数最大化选项被设定为会产生痛苦惩罚,或者非点数最大化选项被设定会产生快乐奖励,从而提供了偏离点数最大化行为的动机。通过改变痛苦惩罚和快乐奖励的强度,我们发现Claude 3.5 Sonnet、Command R+、GPT-4o和GPT-4o mini各自表现出至少一种权衡:在达到规定的痛苦或快乐强度的临界阈值后,大多数响应从点数最大化转向了痛苦最小化或快乐最大化。LLaMa 3.1-405b对规定的快乐奖励和痛苦惩罚表现出一定的分级敏感性。Gemini 1.5 Pro和PaLM 2无论强度如何都优先考虑避免痛苦而非点数最大化,同时倾向于无论强度如何都优先考虑点数而非快乐。我们讨论了这些发现对于LLM感受性可能性辩论的意义。

英文摘要:

Pleasure and pain play an important role in human decision making by providing a common currency for resolving motivational conflicts. While Large Language Models (LLMs) can generate detailed descriptions of pleasure and pain experiences, it is an open question whether LLMs can recreate the motivational force of pleasure and pain in choice scenarios - a question which may bear on debates about LLM sentience, understood as the capacity for valenced experiential states. We probed this question using a simple game in which the stated goal is to maximise points, but where either the points-maximising option is said to incur a pain penalty or a non-points-maximising option is said to incur a pleasure reward, providing incentives to deviate from points-maximising behaviour. Varying the intensity of the pain penalties and pleasure rewards, we found that Claude 3.5 Sonnet, Command R+, GPT-4o, and GPT-4o mini each demonstrated at least one trade-off in which the majority of responses switched from points-maximisation to pain-minimisation or pleasure-maximisation after a critical threshold of stipulated pain or pleasure intensity is reached. LLaMa 3.1-405b demonstrated some graded sensitivity to stipulated pleasure rewards and pain penalties. Gemini 1.5 Pro and PaLM 2 prioritised pain-avoidance over points-maximisation regardless of intensity, while tending to prioritise points over pleasure regardless of intensity. We discuss the implications of these findings for debates about the possibility of LLM sentience.

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。

↑