arXiv cs.CL· Laya Iyer, Sanmi Koyejo, Chen Shani·· 11 小时前AI 评分26
语言模型表征中的情绪压缩现象研究
Compression of Emotion in Language Model Representations
AI 导读
研究在 GoEmotions 数据集上测试七个 LLM，发现其隐藏表征存在"过度压缩"：无监督 KMeans 聚类时最清晰的划分仅将正负情绪分开，粗粒度情绪分组所需的簇数也少于完整的 28 类分类体系。但线性探针仍保留大量细粒度情绪预测信息，说明细粒度情绪信息可被访问，只是无监督几何结构更受粗粒度情感结构主导。
正文 · 原文
Abstract:Large language models (LLMs) have been evaluated on their ability to identify emotions expressed in text and produce fluent emotional language. However, output predictions alone do not reveal how fine-grained emotion distinctions are organized internally. We study seven LLMs on GoEmotions, an English Reddit dataset annotated with 27 fine-grained emotion categories plus neutral. We compare model predictions and emotion-prompted final-token representations with the 28 GoEmotions categories and with coarser emotion-family and sentiment groupings.
We find a consistent pattern of $\textit{overcompression}$ relative to the fine-grained GoEmotions taxonomy. Under the tested KMeans protocol, when we cluster their hidden representations without using emotion labels, the clearest split is into two groups that largely separate positive from negative emotion. Coarser emotion groupings are also recovered with fewer clusters than the full 28-category taxonomy. Yet linear probes retain substantial fine-grained predictive information. These results suggest that fine-grained emotion information remains accessible even though broader affective structure more strongly organizes the unsupervised geometry.
| Subjects: | Computation and Language (cs.CL); Artificial Intelligence (cs.AI) |
| Cite as: | arXiv:2611.08683 [cs.CL] |
| (or arXiv:2611.08683v1 [cs.CL] for this version) | |
| https://doi.org/10.48550/arXiv.2611.08683 arXiv-issued DOI via DataCite (pending registration) |
Submission history
From: Laya Iyer [view email]
[v1]
Thu, 5 Nov 2026 20:42:47 UTC (1,271 KB)
来源：arXiv cs.CL · arxiv.org