arXiv cs.CL· Joseph Marks (NatWest Group, London, UK), Marius C. Vollberg (University of Zurich, Zurich, Switzerland, Swiss Center for Affective Sciences, Geneva, Switzerland)·· 11 小时前AI 评分31
大语言模型中的情感心理物理学研究
The psychophysics of affect in large language models
AI 导读
研究测试了 LLM 如何预判赌博或股票交易结果带来的感受，以及是否将评分纳入选择。单独评分时，模型对量级的敏感度递减、类似人类价值函数，但对收益的评分强度高于等额损失，形成"硅基收益放大"这一对经典损失厌恶的反转；合并评分时评分线性化，选择中的得失不对称基本消失，模型自身评分并不比金钱效用更好地解释其选择。
正文 · 原文
Authors:Joseph Marks (1), Marius C. Vollberg (2 and 3) ((1) NatWest Group, London, UK, (2) University of Zurich, Zurich, Switzerland, (3) Swiss Center for Affective Sciences, Geneva, Switzerland)
Abstract:Good decisions often require considering how outcomes will feel. Yet large language models (LLMs) increasingly enter decision workflows, despite little being known about how they map outcomes to feelings or use those mappings in choice. We tested how LLMs anticipated gambling or stock-trading outcomes would feel and whether they incorporated these ratings into choice. When outcomes were rated in isolation, ratings exhibited diminishing sensitivity to magnitude, resembling human value functions; but models rated gains more intensely than equivalent losses, an inversion of canonical loss aversion we call silicon gain amplification. When outcomes were rated together, enabling direct comparison, ratings linearised. In choice, the gain-loss asymmetry largely disappeared, and models' own ratings did not reliably explain their choices better than monetary utility did. These findings reveal malleable mappings from outcomes to affect, raising questions about whether and how agentic LLMs should incorporate anticipated feelings into decisions.
| Comments: | 64 pages (26 main text, 25 Supplementary Information, 13 GUIDE-LLM reporting checklist), 5 figures. Code: this https URL ; data: this https URL |
| Subjects: | Artificial Intelligence (cs.AI); Computation and Language (cs.CL) |
| ACM classes: | I.2.7; J.4 |
| Cite as: | arXiv:2610.40985 [cs.AI] |
| (or arXiv:2610.40985v1 [cs.AI] for this version) | |
| https://doi.org/10.48550/arXiv.2610.40985 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Joseph Marks [view email]
[v1]
Thu, 29 Oct 2026 21:09:12 UTC (1,491 KB)
来源：arXiv cs.CL · arxiv.org