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.
Increasing happiness through conversations with artificial intelligence
Yale University(耶鲁大学);Sainsbury Wellcome Centre for Neural Circuits and Behaviour, UCL(伦敦大学学院Sainsbury Wellcome神经环路与行为中心);Google DeepMind(谷歌DeepMind);University of Oxford(牛津大学);Max Planck UCL Centre for Computational Psychiatry and Ageing Research(马克斯·普朗克伦敦大学学院计算精神病学与衰老研究中心)
EmotionQueen: A Benchmark for Evaluating Empathy of Large Language Models
Fudan University(复旦大学);School of Computer Science, Fudan University(复旦大学计算机科学学院);Beijing Yuandian Technology(北京原点科技);School of Engineering and Applied Science, University of Pennsylvania(宾夕法尼亚大学工程与应用科学学院)
How Emotion Shapes the Behavior of LLMs and Agents: A Mechanistic Study
School of Computer Science and Engineering, Beihang University(北京航空航天大学计算机科学与工程学院);College of Computing and Data Science, Nanyang Technological University(南洋理工大学计算与数据科学学院)
Information Technologies Group, atlanTTic, University of Vigo, Vigo, Spain(信息科技组,atlanTTic,维戈大学);Research on Economics, Management and Information Technologies, Universidade Portucalense, Porto, Portugal(经济学、管理与信息科技研究,葡萄牙波尔图大学);ISEP, Polytechnic of Porto, Porto, Portugal(ISEP,波尔图理工学院);Institute for Systems and Computer Engineering, Technology and Science, Porto, Portugal(系统与计算机工程、技术与科学研究所)
LLMs Capture Emotion Labels, Not Emotion Uncertainty: Distributional Analysis and Calibration of Human-LLM Judgment Gaps
Faculty of Business and Commerce(商科学院);Faculty of Business Data Science(商务数据科学学院);RIKEN Center for Advanced Intelligence Project(先进智能项目研究所);Faculty of Data Science(数据科学学院);Japan Safety Society Research Center(日本安全协会研究中心)
CULEMO: Cultural Lenses on Emotion -- Benchmarking LLMs for Cross-Cultural Emotion Understanding
Instituto Politécnico Nacional(墨西哥国立理工学院);Haverford College(哈弗福德学院);Wollo University(沃尔洛大学);Pennsylvania State University(宾夕法尼亚州立大学);University of Hamburg(汉堡大学)
Large Language Models Understand and Can be Enhanced by Emotional Stimuli
Institute of Software, CAS(中国科学院软件研究所);Microsoft(微软);William&Mary(威廉与玛丽学院);Department of Psychology, Beijing Normal University(北京师范大学心理学系);HKUST(香港科技大学)
CAREBench: Evaluating LLMs' Emotion Understanding by Assessing Cognitive Appraisal Reasoning
Department of Informatics(信息学院);King’s College London(伦敦国王学院);Institute of Psychology(心理学研究所);University of Tartu(塔尔图大学);Department of Psychology(心理学系);Stanford University(斯坦福大学);The Alan Turing Institute(艾伦·图灵研究所)
AI with Emotions: Exploring Emotional Expressions in Large Language Models
Graduate School of Artificial Intelligence and Science, Rikkyo University(人工智能与科学研究生院,立命馆大学);Strategic Digital Business Unit, Mamezou Co., Ltd.(战略数字商务部,麦默株式会社)
DialogueLLM: Context and Emotion Knowledge-Tuned Large Language Models for Emotion Recognition in Conversations
The Hong Kong Polytechnic University(香港理工大学);Zhengzhou University of Light Industry(郑州轻工业大学);Hebei University of Technology(河北工业大学);Halmstad University(哈尔姆斯塔德大学);Copenhagen University(哥本哈根大学);The Chinese University of Hong Kong, Shenzhen(香港中文大学(深圳))
VoxEmo: Benchmarking Speech Emotion Recognition with Speech LLMs
Department of Computer Science, University of Sheffield, United Kingdom(英国谢菲尔德大学计算机科学系);Signal Analysis and Interpretation Laboratory (SAIL), Ming Hsieh Department of Electrical and Computer Engineering, University of Southern California, Los Angeles, CA 90089, USA(美国南加州大学电气与计算机工程系信号分析与解释实验室(SAIL))
Effectiveness of Large Language Models in Simulating Regional Psychological Structures: An Empirical Examination of Personality and Subjective Well-being
Tsinghua University(清华大学);Neoma Business School(诺欧商学院);Beijing Technology and Business University(北京工商大学);Beijing Normal University(北京师范大学)
How People Respond to the COVID-19 Pandemic on Twitter: A Comparative Analysis of Emotional Expressions from US and India
Institute of High Performance Computing (IHPC), Agency for Science, Technology and Research (A*STAR)(高性能计算研究所,新加坡科技研究局);Centre for Artificial Intelligence Research, University of Agder(阿格德大学人工智能研究中心);Department of Psychology, Northwestern University(西北大学心理学系)