---
title: "Large Language Models are Superpositions of All Characters:"
description: "They propose DITTO, a self-alignment recipe that needs no external teacher. 23 references, cited by 63 papers here."
authors: ["Keming Lu", "Bowen Yu", "Chang Zhou", "Jingren Zhou"]
year: 2024
venue: "Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)"
source: "https://aclanthology.org/2024.acl-long.423/"
doi: "10.18653/v1/2024.acl-long.423"
licence: "CC BY 4.0 (ACL Anthology)"
licence_url: "https://creativecommons.org/licenses/by/4.0/"
summary_licence: "Elamur Cosmic License"
canonical: "https://research.elamur.ai/free/en/papers/lu-2024-large-language-models-are-superpositions/"
---
[Elamur Research](https://md.research.elamur.ai/free/en/) · [this page in HTML](https://research.elamur.ai/free/en/papers/lu-2024-large-language-models-are-superpositions/)

[home](https://md.research.elamur.ai/free/en/) / [papers](https://md.research.elamur.ai/free/en/papers/) / [2024](https://md.research.elamur.ai/free/en/papers/2024/page/4/) / Large Language Models are Superpositions of All Characters: Attaining Arbitrary Role-play via Self-Alignment

# Large Language Models are Superpositions of All Characters: Attaining Arbitrary Role-play via Self-Alignment

**Authors:** [Keming Lu](https://md.research.elamur.ai/free/en/authors/keming-lu/), [Bowen Yu](https://md.research.elamur.ai/free/en/authors/bowen-yu/), [Chang Zhou](https://md.research.elamur.ai/free/en/authors/chang-zhou/) and [Jingren Zhou](https://md.research.elamur.ai/free/en/authors/jingren-zhou/) · **Year:** 2024 · **Venue:** [Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)](https://md.research.elamur.ai/free/en/venues/acl/) ([ACL 2024](https://md.research.elamur.ai/free/en/venues/acl/2024/)) · **Publisher:** [Association for Computational Linguistics](https://md.research.elamur.ai/free/en/publishers/association-for-computational-linguistics/) · **Licence:** [CC BY 4.0 (ACL Anthology)](https://creativecommons.org/licenses/by/4.0/)

**Links:** [Publisher's page](https://aclanthology.org/2024.acl-long.423/) · [PDF](https://aclanthology.org/2024.acl-long.423.pdf) · [DOI 10.18653/v1/2024.acl-long.423](https://doi.org/10.18653/v1/2024.acl-long.423) · [OpenAlex](https://openalex.org/W4402671594)

**Citations:** 23 citations (OpenAlex, 2026-10-14) · [63](https://md.research.elamur.ai/free/en/papers/lu-2024-large-language-models-are-superpositions/cited-by/) in our library, of the 141,840 papers whose bibliographies we read

## Summary
Open-source role-play chatbots are usually trained on dialogues distilled from proprietary systems such as GPT-4. This requires a stronger teacher, tends to increase hallucination, and brings licensing constraints. The authors argue that an LLM already holds knowledge of countless characters from pre-training. They propose DITTO, a self-alignment recipe that needs no external teacher.

DITTO works in two stages:
- **Simulation:** an instruction-following model receives a character's attributes and profile collected from Wikipedia. It generates user queries and in-character replies, treating the task like reading comprehension.
- **Fine-tuning:** the model is trained to give those same replies when shown only a short character description. This forces it to draw on its internal knowledge.

The process yields WikiRole, a multilingual self-generated dataset covering 4,000 characters. The authors also build an automatic, LLM-judged benchmark that checks identity consistency, role-specific factual accuracy, and refusal of out-of-scope questions.

Applied to Qwen-Chat models, DITTO beats all open-source role-play baselines and approaches proprietary chatbots. Cross-supervision experiments show that conversational style transfers easily even from smaller models, whereas role knowledge is limited by the base model itself.

## Key points
- WikiRole covers 4,000 Wikipedia characters, about ten times more roles than earlier public role-play datasets. Resources are released at github.com/OFA-Sys/Ditto.
- DITTO was applied to four Qwen-Chat sizes (1.8B, 7B, 14B, 72B). The 72B version reaches about 90% on role identity consistency and is roughly on par with GPT-3.5-Turbo. It still trails GPT-4 and Qwen-Max on role-related knowledge.
- The benchmark reduces evaluation to judge-answerable questions:
  - choosing the portrayed character from four options;
  - checking the dialogue against gold knowledge;
  - testing whether the model admits ignorance, e.g., when Harry Potter is asked about quicksort.
- Evaluation also uses the role-play subset of MT-Bench.
- In cross-supervision, identity consistency improves after imitation even when the teacher is weaker. Knowledge and rejection scores, by contrast, are capped by the seed model.
- Weak-to-strong generalization appears on knowledge metrics. Qwen-72B-Chat tuned on simulations from Qwen-1.8B-Chat matches self-aligned Qwen-7B-Chat.

*Our summary of the paper, in our own words — not its text. The work belongs to its authors; read it at the links above.*

## Institutions

- [Alibaba Group](https://md.research.elamur.ai/free/en/companies/alibaba-group/) — Company, Hangzhou, China · 64 works · 14,113 citations

📚 Shelves (4): [🖥️ Computer Science](https://md.research.elamur.ai/free/en/shelves/computer-science/) · 9,086 works · [🤖 Artificial Intelligence](https://md.research.elamur.ai/free/en/shelves/computer-science/artificial-intelligence/) · 8,889 works · [🎭 Arts and Humanities](https://md.research.elamur.ai/free/en/shelves/arts-and-humanities/) · 9,296 works · [🗣️ Language and Linguistics](https://md.research.elamur.ai/free/en/shelves/arts-and-humanities/language-and-linguistics/) · 9,206 works

🧭 Topics (1): [🧠 Language models & pretraining](https://md.research.elamur.ai/free/en/topics/language-models/) · 1,343 works

## Cite

Cite as: Keming Lu, Bowen Yu, Chang Zhou and Jingren Zhou. 2024. Large Language Models are Superpositions of All Characters: Attaining Arbitrary Role-play via Self-Alignment. Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers).

The paper itself, not this page. BibTeX from the library's record of the work.

```bibtex
@inproceedings{lu2024large,
  title = {{Large Language Models are Superpositions of All Characters: Attaining Arbitrary Role-play via Self-Alignment}},
  author = {Keming Lu, Bowen Yu, Chang Zhou and Jingren Zhou},
  booktitle = {Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)},
  year = {2024},
  doi = {10.18653/v1/2024.acl-long.423},
  url = {https://doi.org/10.18653/v1/2024.acl-long.423}
}
```

- [← previous Self-Checker: Plug-and-Play Modules for Fact-Checking with Large Language Models](https://md.research.elamur.ai/free/en/papers/li-2024-self-checker-plug-play-modules/)
- [next → LLM Comparative Assessment: Zero-shot NLG Evaluation through Pairwise Comparisons using Large Language Models](https://md.research.elamur.ai/free/en/papers/liusie-2024-llm-comparative-assessment-zero-shot/)
- [↑ back to papers of 2024](https://md.research.elamur.ai/free/en/papers/2024/page/4/)

## References

23 references, 13 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

1. Jinze Bai, Shuai Bai, Yunfei Chu, Zeyu Cui, Kai Dang, Xiaodong Deng, Yang Fan, Wenbin Ge, Yu Han, Fei Huang, et al. 2023. [Qwen technical report](https://md.research.elamur.ai/free/en/papers/bai-2023-qwen-technical-report/). arXiv preprint arXiv:2309.16609.
2. Tom Brown, Benjamin Mann, Nick Ryder, Melanie Subbiah, Jared D Kaplan, Prafulla Dhariwal, Arvind Neelakantan, Pranav Shyam, Girish Sastry, Amanda Askell, et al. 2020. [Language models are few-shot learners](https://md.research.elamur.ai/free/en/papers/brown-2020-language-models-few-shot-learners/). Advances in neural information processing systems, 33:1877–1901.
3. Collin Burns, Pavel Izmailov, Jan Hendrik Kirchner, Bowen Baker, Leo Gao, Leopold Aschenbrenner, Yining Chen, Adrien Ecoffet, Manas Joglekar, Jan Leike, Ilya Sutskever, and Jeff Wu. 2023. [Weak-to-strong generalization: Eliciting strong capabilities with weak supervision](https://md.research.elamur.ai/free/en/papers/burns-2023-weak-strong-generalization-eliciting-strong/).
4. Nuo Chen, Yan Wang, Haiyun Jiang, Deng Cai, Yuhan Li, Ziyang Chen, Longyue Wang, and Jia Li. 2023. [Large language models meet harry potter: A dataset for aligning dialogue agents with characters](https://md.research.elamur.ai/free/en/papers/chen-2023-large-language-models-meet-harry/). In Findings of the Association for Computational Linguistics: EMNLP 2023, pages 8506–8520.
5. Keming Lu, Hongyi Yuan, Zheng Yuan, Runji Lin, Junyang Lin, Chuanqi Tan, Chang Zhou, and Jingren Zhou. 2023b. \# InsTag: Instruction tagging for analyzing supervised fine-tuning of large language models. arXiv e-prints, pages arXiv–2308. [find](https://scholar.google.com/scholar?q=Keming+Lu%2C+Hongyi+Yuan%2C+Zheng+Yuan%2C+Runji+Lin%2C+Junyang+Lin%2C+Chuanqi+Tan%2C+Chang+Zhou%2C+and+Jingren+Zhou%2E+2023b%2E+%5C%23+InsTag%3A+Instruction+tagging+for+analyzing+supervised+fine%2Dtuning+of+large+language+mode)
6. Niklas Muennighoff, Qian Liu, Armel Zebaze, Qinkai Zheng, Binyuan Hui, Terry Yue Zhuo, Swayam Singh, Xiangru Tang, Leandro Von Werra, and Shayne Longpre. 2023. Octopack: Instruction tuning code large language models. arXiv preprint arXiv:2308.07124. [find](https://scholar.google.com/scholar?q=Niklas+Muennighoff%2C+Qian+Liu%2C+Armel+Zebaze%2C+Qinkai+Zheng%2C+Binyuan+Hui%2C+Terry+Yue+Zhuo%2C+Swayam+Singh%2C+Xiangru+Tang%2C+Leandro+Von+Werra%2C+and+Shayne+Longpre%2E+2023%2E+Octopack%3A+Instruction+tuning+code+large+)
7. Murray Shanahan, Kyle McDonell, and Laria Reynolds. 2023. [Role play with large language models](https://md.research.elamur.ai/free/en/papers/shanahan-2023-role-play-large-language-models/). Nature, pages 1–6.
8. Yunfan Shao, Linyang Li, Junqi Dai, and Xipeng Qiu. 2023. [Character-llm: A trainable agent for role-playing](https://md.research.elamur.ai/free/en/papers/shao-2023-character-llm-trainable-agent-role/). arXiv preprint arXiv:2310.10158.
9. Meiling Tao, Xuechen Liang, Tianyu Shi, Lei Yu, and Yiting Xie. 2023. Rolecraft-glm: Advancing personalized role-playing in large language models. [find](https://scholar.google.com/scholar?q=Meiling+Tao%2C+Xuechen+Liang%2C+Tianyu+Shi%2C+Lei+Yu%2C+and+Yiting+Xie%2E+2023%2E+Rolecraft%2Dglm%3A+Advancing+personalized+role%2Dplaying+in+large+language+models%2E)
10. Arnav Gudibande, Eric Wallace, Charlie Snell, Xinyang Geng, Hao Liu, Pieter Abbeel, Sergey Levine, and Dawn Song. 2023. [The false promise of imitating proprietary llms](https://md.research.elamur.ai/free/en/papers/gudibande-2023-false-promise-imitating-proprietary-llms/). arXiv preprint arXiv:2305.15717.
11. Hugo Touvron, Louis Martin, Kevin Stone, Peter Albert, Amjad Almahairi, Yasmine Babaei, Nikolay Bashlykov, Soumya Batra, Prajjwal Bhargava, Shruti Bhosale, et al. 2023. Llama 2: Open foundation and fine-tuned chat models. arXiv preprint arXiv:2307.09288. [find](https://scholar.google.com/scholar?q=Hugo+Touvron%2C+Louis+Martin%2C+Kevin+Stone%2C+Peter+Albert%2C+Amjad+Almahairi%2C+Yasmine+Babaei%2C+Nikolay+Bashlykov%2C+Soumya+Batra%2C+Prajjwal+Bhargava%2C+Shruti+Bhosale%2C+et+al%2E+2023%2E+Llama+2%3A+Open+foundation+and+fi)
12. Albert Q Jiang, Alexandre Sablayrolles, Arthur Mensch, Chris Bamford, Devendra Singh Chaplot, Diego de las Casas, Florian Bressand, Gianna Lengyel, Guillaume Lample, Lucile Saulnier, et al. 2023. Mistral 7b. arXiv preprint arXiv:2310.06825. [find](https://scholar.google.com/scholar?q=Albert+Q+Jiang%2C+Alexandre+Sablayrolles%2C+Arthur+Mensch%2C+Chris+Bamford%2C+Devendra+Singh+Chaplot%2C+Diego+de+las+Casas%2C+Florian+Bressand%2C+Gianna+Lengyel%2C+Guillaume+Lample%2C+Lucile+Saulnier%2C+et+al%2E+2023%2E+Mist)
13. Quan Tu, Chuanqi Chen, Jinpeng Li, Yanran Li, Shuo Shang, Dongyan Zhao, Ran Wang, and Rui Yan. 2023. Characterchat: Learning towards conversational ai with personalized social support. arXiv preprint arXiv:2308.10278. — [arXiv:2308.10278](https://arxiv.org/abs/2308.10278) [find](https://scholar.google.com/scholar?q=Quan+Tu%2C+Chuanqi+Chen%2C+Jinpeng+Li%2C+Yanran+Li%2C+Shuo+Shang%2C+Dongyan+Zhao%2C+Ran+Wang%2C+and+Rui+Yan%2E+2023%2E+Characterchat%3A+Learning+towards+conversational+ai+with+personalized+social+support%2E+arXiv+preprint+)
14. Albert Q. Jiang, Alexandre Sablayrolles, Antoine Roux, Arthur Mensch, Blanche Savary, Chris Bamford, Devendra Singh Chaplot, Diego de las Casas, Emma Bou Hanna, Florian Bressand, Gianna Lengyel, Guillaume Bour, Guillaume Lample, Lélio Renard Lavaud, Lucile Saulnier, Marie-Anne Lachaux, Pierre Stock,… [find](https://scholar.google.com/scholar?q=Albert+Q%2E+Jiang%2C+Alexandre+Sablayrolles%2C+Antoine+Roux%2C+Arthur+Mensch%2C+Blanche+Savary%2C+Chris+Bamford%2C+Devendra+Singh+Chaplot%2C+Diego+de+las+Casas%2C+Emma+Bou+Hanna%2C+Florian+Bressand%2C+Gianna+Lengyel%2C+Guill)
15. Guan Wang, Sijie Cheng, Xianyuan Zhan, Xiangang Li, Sen Song, and Yang Liu. 2023a. [Openchat: Advancing open-source language models with mixed-quality data](https://md.research.elamur.ai/free/en/papers/wang-2023-openchat-advancing-open-source-language/). arXiv preprint arXiv:2309.11235.
16. Yufei Wang, Wanjun Zhong, Liangyou Li, Fei Mi, Xingshan Zeng, Wenyong Huang, Lifeng Shang, Xin Jiang, and Qun Liu. 2023b. [Aligning large language models with human: A survey](https://md.research.elamur.ai/free/en/papers/wang-2023-aligning-large-language-models-human/). arXiv preprint arXiv:2307.12966.
17. Cheng Li, Ziang Leng, Chenxi Yan, Junyi Shen, Hao Wang, Weishi MI, Yaying Fei, Xiaoyang Feng, Song Yan, HaoSheng Wang, et al. 2023a. [Chatharuhi: Reviving anime character in reality via large language model](https://md.research.elamur.ai/free/en/papers/li-2023-chatharuhi-reviving-anime-character-reality/). arXiv preprint arXiv:2308.09597.
18. Zekun Moore Wang, Zhongyuan Peng, Haoran Que, Jiaheng Liu, Wangchunshu Zhou, Yuhan Wu, Hongcheng Guo, Ruitong Gan, Zehao Ni, Man Zhang, et al. 2023c. [Rolellm: Benchmarking, eliciting, and enhancing role-playing abilities of large language models](https://md.research.elamur.ai/free/en/papers/wang-2023-rolellm-benchmarking-eliciting-enhancing-role/). arXiv preprint arXiv:2310.00746.
19. Xian Li, Ping Yu, Chunting Zhou, Timo Schick, Luke Zettlemoyer, Omer Levy, Jason Weston, and Mike Lewis. 2023b. [Self-alignment with instruction back-translation](https://md.research.elamur.ai/free/en/papers/li-2023-self-alignment-instruction-backtranslation/). arXiv preprint arXiv:2308.06259.
20. Mengge Xue, Bowen Yu, Zhenyu Zhang, Tingwen Liu, Yue Zhang, and Bin Wang. 2020. Coarse-to-fine pre-training for named entity recognition. arXiv preprint arXiv:2010.08210. [find](https://scholar.google.com/scholar?q=Mengge+Xue%2C+Bowen+Yu%2C+Zhenyu+Zhang%2C+Tingwen+Liu%2C+Yue+Zhang%2C+and+Bin+Wang%2E+2020%2E+Coarse%2Dto%2Dfine+pre%2Dtraining+for+named+entity+recognition%2E+arXiv+preprint+arXiv%3A2010%2E08210%2E)
21. Keming Lu, Xiaoman Pan, Kaiqiang Song, Hongming Zhang, Dong Yu, and Jianshu Chen. 2023a. Pivoine: Instruction tuning for open-world entity profiling. In Findings of the Association for Computational Linguistics: EMNLP 2023, pages 15108–15127. [find](https://scholar.google.com/scholar?q=Keming+Lu%2C+Xiaoman+Pan%2C+Kaiqiang+Song%2C+Hongming+Zhang%2C+Dong+Yu%2C+and+Jianshu+Chen%2E+2023a%2E+Pivoine%3A+Instruction+tuning+for+open%2Dworld+entity+profiling%2E+In+Findings+of+the+Association+for+Computational+L)
22. Jifan Yu, Xiaohan Zhang, Yifan Xu, Xuanyu Lei, Xinyu Guan, Jing Zhang, Lei Hou, Juanzi Li, and Jie Tang. 2022. Xdai: A tuning-free framework for exploiting pre-trained language models in knowledge grounded dialogue generation. In Proceedings of the 28th ACM SIGKDD Conference on Knowledge Discovery a… [find](https://scholar.google.com/scholar?q=Jifan+Yu%2C+Xiaohan+Zhang%2C+Yifan+Xu%2C+Xuanyu+Lei%2C+Xinyu+Guan%2C+Jing+Zhang%2C+Lei+Hou%2C+Juanzi+Li%2C+and+Jie+Tang%2E+2022%2E+Xdai%3A+A+tuning%2Dfree+framework+for+exploiting+pre%2Dtrained+language+models+in+knowledge+gro)
23. Query for collecting virtual characters Zheng Yuan, Hongyi Yuan, Chengpeng Li, Guanting Dong, Keming Lu, Chuanqi Tan, Chang Zhou, and Jingren Zhou. 2023. [Scaling relationship on learning mathematical reasoning with large language models](https://md.research.elamur.ai/free/en/papers/yuan-2023-scaling-relationship-learning-mathematical-reasoning/).

*Updated 24 October 2026*