Saved in:
| Main Authors: | , , , , , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2504.10227 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866918106220199936 |
|---|---|
| author | Ju, Tianjie Shao, Zhenyu Wang, Bowen Chen, Yujia Zhang, Zhuosheng Fei, Hao Lee, Mong-Li Hsu, Wynne Duan, Sufeng Liu, Gongshen |
| author_facet | Ju, Tianjie Shao, Zhenyu Wang, Bowen Chen, Yujia Zhang, Zhuosheng Fei, Hao Lee, Mong-Li Hsu, Wynne Duan, Sufeng Liu, Gongshen |
| contents | Large Language Models (LLMs) have demonstrated promising capabilities to generate responses that simulate consistent personality traits. Despite the major attempts to analyze personality expression through output-based evaluations, little is known about how such traits are internally encoded within LLM parameters. In this paper, we introduce a layer-wise probing framework to systematically investigate the layer-wise capability of LLMs in simulating personality for responding. We conduct probing experiments on 11 open-source LLMs over the PersonalityEdit benchmark and find that LLMs predominantly simulate personality for responding in their middle and upper layers, with instruction-tuned models demonstrating a slightly clearer separation of personality traits. Furthermore, by interpreting the trained probing hyperplane as a layer-wise boundary for each personality category, we propose a layer-wise perturbation method to edit the personality expressed by LLMs during inference. Our results show that even when the prompt explicitly specifies a particular personality, our method can still successfully alter the response personality of LLMs. Interestingly, the difficulty of converting between certain personality traits varies substantially, which aligns with the representational distances in our probing experiments. Finally, we conduct a comprehensive MMLU benchmark evaluation and time overhead analysis, demonstrating that our proposed personality editing method incurs only minimal degradation in general capabilities while maintaining low training costs and acceptable inference latency. Our code is publicly available at https://github.com/universe-sky/probing-then-editing-personality. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2504_10227 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | Probing then Editing Response Personality of Large Language Models Ju, Tianjie Shao, Zhenyu Wang, Bowen Chen, Yujia Zhang, Zhuosheng Fei, Hao Lee, Mong-Li Hsu, Wynne Duan, Sufeng Liu, Gongshen Computation and Language Large Language Models (LLMs) have demonstrated promising capabilities to generate responses that simulate consistent personality traits. Despite the major attempts to analyze personality expression through output-based evaluations, little is known about how such traits are internally encoded within LLM parameters. In this paper, we introduce a layer-wise probing framework to systematically investigate the layer-wise capability of LLMs in simulating personality for responding. We conduct probing experiments on 11 open-source LLMs over the PersonalityEdit benchmark and find that LLMs predominantly simulate personality for responding in their middle and upper layers, with instruction-tuned models demonstrating a slightly clearer separation of personality traits. Furthermore, by interpreting the trained probing hyperplane as a layer-wise boundary for each personality category, we propose a layer-wise perturbation method to edit the personality expressed by LLMs during inference. Our results show that even when the prompt explicitly specifies a particular personality, our method can still successfully alter the response personality of LLMs. Interestingly, the difficulty of converting between certain personality traits varies substantially, which aligns with the representational distances in our probing experiments. Finally, we conduct a comprehensive MMLU benchmark evaluation and time overhead analysis, demonstrating that our proposed personality editing method incurs only minimal degradation in general capabilities while maintaining low training costs and acceptable inference latency. Our code is publicly available at https://github.com/universe-sky/probing-then-editing-personality. |
| title | Probing then Editing Response Personality of Large Language Models |
| topic | Computation and Language |
| url | https://arxiv.org/abs/2504.10227 |