发现开源大模型中存在情绪向量，且其分布模式各异。
Where Do Models Find Happiness? Emotion Vectors in Open-Source LLMs
- 通过对比向量提取，分析模型各层的情绪表征结构。
- 两个模型均呈现显著情感维度相关性，最高达r=0.83。
- 情绪编码方式因模型和生成语料而异，适合研究模型内部心理机制者阅读。
近期研究在Claude Sonnet 4.5中发现了情绪向量，这些是编码情绪概念的内部表征，能因果影响行为，并具有与人类心理结构一致的几何特征。本文在两个开源模型Apertus-8B-Instruct-2509和Gemma-4-E4B-it中验证该现象，跨所有层提取情绪对比向量，并使用两个由模型生成的语料库进行测试。结果发现两模型均恢复出效价（valence）几何结构，峰值主成分1（PC1）与效价的相关系数分别为r = 0.76和r = 0.83，接近Claude报告的r = 0.81。超越复现，我们观察到效价表征在模型深度上呈现显著差异：Gemma-4-E4B-it中效价在早期层强编码，后期逐渐衰减；而Apertus-8B-Instruct-2509则相反，早期无效价信号，中层才出现。唤醒度（arousal）编码则对数据集敏感：两模型在Gemma生成的故事中表现出更强的PC2–唤醒度对齐（r最高达0.45），而在Apertus生成语料中对齐较弱（r ≤ 0.21），表明唤醒度相关线索在生成语料中分布不均。研究已开源实验代码与数据集，支持对不同架构下情绪表征的可复现研究。
原文摘要 · Abstract (English)
Recent work identified emotion vectors in Claude Sonnet 4.5, which are internal representations that encode emotion concepts, causally influence behavior, and exhibit geometry mirroring human psychological structure. We test the generality of these findings in two open-weight models, Apertus-8B-Instruct-2509 and Gemma-4-E4B-it, extracting emotion contrast vectors across all layers, using two model-generated corpora. We recover valence geometry for both models, with peak PC1--valence correlations of $r = 0.76$ and $r = 0.83$, approaching the $r = 0.81$ reported for Claude.Beyond replication, we observe notable differences in how valence representations emerge across model depth. In Gemma-4-E4B-it, valence is strongly encoded in early layers but collapses towards later layers, whereas Apertus-8B-Instruct-2509 exhibits the opposite pattern, with valence representations absent in early layers, but emerging at mid depths. Arousal encoding, in contrast, is sensitive to the extraction corpus: both models show stronger PC2--arousal alignment with Gemma-generated stories ($r$ up to $0.45$) than Apertus-generated ones ($r \leq 0.21$), suggesting arousal-relevant cues are unevenly distributed across generated corpora. We open-source our experiment code and dataset for reproducible investigation of emotion representations across language model architectures.
Thank you to arXiv for use of its open access interoperability. PaperDance 不是 arXiv 官方产品；中文卡片由大模型生成，请以原文为准。