Language-specific representation of emotion-concept knowledge causally supports emotion inference
Humans no doubt use language to communicate about their emotional experiences, but does language in turn help humans understand emotions, or is language just a vehicle of communication? This study used a form of artificial intelligence (AI) known as large language models (LLMs) to assess whether language-based representations of emotion causally contribute to the AI's ability to generate inferences about the emotional meaning of novel situations. Fourteen attributes of human emotion concept representation were found to be represented by the LLM's distinct artificial neuron populations. By manipulating these attribute-related neurons, we in turn demonstrated the role of emotion concept knowledge in generative emotion inference. The attribute-specific performance deterioration was related to the importance of different attributes in human mental space. Our findings provide a proof-in-concept that even an LLM can learn about emotions in the absence of sensory-motor representations and highlight the contribution of language-derived emotion-concept knowledge for emotion inference.
Authors
- Yu-Sheng Su (ORCID: https://orcid.org/0000-0002-1531-3363)
- Jiali Cheng (ORCID: https://orcid.org/0000-0003-4548-9295)
- Kristen A. Lindquist (ORCID: https://orcid.org/0000-0002-5368-9302)
- Hsiu-Yuan Huang
- Zhiyuan Liu (ORCID: https://orcid.org/0000-0002-7709-2543)
- Ming Li (ORCID: https://orcid.org/0000-0001-5948-1054)
- Dan Zhang (ORCID: https://orcid.org/0000-0002-7592-3200)
- Xin Hu (ORCID: https://orcid.org/0009-0001-9117-8300)
- Huadong Wang (ORCID: https://orcid.org/0000-0002-2803-4814)
- Xiaozhi Wang (ORCID: https://orcid.org/0000-0002-9829-9332)
- Xinmiao Zhang (ORCID: https://orcid.org/0000-0002-1622-1452)
- Yujia Qin
Institutions
- University of North Carolina at Chapel Hill (US)
- University of Massachusetts Lowell (US)
- University of Pittsburgh (US)
- University of Science and Technology Beijing (CN)
- Tsinghua University (CN)
Publication Details
- Journal
- iScience
- Published
- 2024-11-15
- DOI
- https://doi.org/10.1016/j.isci.2024.111401
- Citations
- 5
- Primary Topic
- Emotions and Moral Behavior
- Type
- article
- Field-Weighted Citation Impact
- 3.22
Funders
- National Science Foundation
- Deutsche Forschungsgemeinschaft
- National Natural Science Foundation of China
- Ministry of Education of the People's Republic of China
- Tsinghua University
- National Key Research and Development Program of China