Elamur Research⌕
☰
home / papers / 2026 / ToMELP: A Theory-of-Mind Benchmark for Route-Controlled Persuasion under the Elaboration Likelihood Model

ToMELP: A Theory-of-Mind Benchmark for Route-Controlled Persuasion under the Elaboration Likelihood Model

Paper · 2026 · Ruirui Wang, Haoran Zhang, Tian Lan et al.

Summary

ToMELP is a benchmark that tests whether large language models can infer a listener's mental state during a persuasive conversation and then use that inference to shape what they say. It also checks whether the output stays within a specified Elaboration Likelihood Model (ELM) route. The authors extend the dialogue-state backbone of PersuasiveToM. Each state contains:

  • the conversation so far,
  • a set of candidate strategies,
  • a target strategy with an effectiveness label,
  • belief/desire/intent (BDI) annotations for both the persuader and the persuadee.

Every state appears in two controlled versions. The central version supplies citable facts and counterpoints. The peripheral version supplies heuristic cue phrases. Crossed with persona, this gives a 2 × 2 design. Scoring is done by an LLM judge and covers attitude change, route compliance, evidence quality on the central route, and robustness to perturbations.

The main findings:

  • Injecting ToM gives steadier gains on the peripheral route.
  • Central-route success depends mostly on how well the model builds arguments.
  • Better ToM alignment does not guarantee route fidelity; models sometimes replace argumentation with cue-based tactics.

The work argues for evaluating personalized persuasion jointly on outcomes, route adherence, and evidence auditability.

Key points

  • The benchmark builds on the public PersuasiveToM dataset. It adds route-specific fields: central_evidence (e.g., facts F1–F6, counterpoints C1–C3) and heuristic_phrases covering authority, social proof, liking, reciprocity and scarcity.
  • Candidate route resources were produced by GPT-4o. Nine in-house research assistants (CS master's students) then checked and revised them item by item.
  • Evaluation relies only on LLM-as-a-judge with structured rubrics. The authors state explicitly that no human agreement validation was performed.
  • Under conflicting, noisy or misleading perturbations, ToM injection behaves like a "goal anchor" and improves overall robustness.
  • Acknowledged limitations:
    • only two persona types and two routes;
    • fixed strategy sets;
    • discrete BDI labels;
    • possible distribution bias from how the resources were constructed.

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

 Fields, topics and concepts

📚 Shelves 6The field of science this work belongs to (OpenAlex fields).Computer Science · 16,667 worksArtificial Intelligence · 16,194 worksArts and Humanities · 16,039 worksLanguage and Linguistics · 15,891 worksSocial Sciences · 1,772 worksSociology and Political Science · 577 works

🧭 Topics 1Research subjects and languages the work is about.Evaluation & benchmarks · 1,220 works

Cite

Cite as Ruirui Wang, Haoran Zhang, Tian Lan, Zehua Duo, Jiang Li, Guanglai Gao and Xiangdong Su. 2026. ToMELP: A Theory-of-Mind Benchmark for Route-Controlled Persuasion under the Elaboration Likelihood Model. Findings of the Association for Computational Linguistics: ACL 2026.

The paper itself, not this page. BibTeX from the library's record of the work. DOI 10.18653/v1/2026.findings-acl.1072

@inproceedings{wang2026tomelp,
  title = {{ToMELP: A Theory-of-Mind Benchmark for Route-Controlled Persuasion under the Elaboration Likelihood Model}},
  author = {Ruirui Wang, Haoran Zhang, Tian Lan, Zehua Duo, Jiang Li, Guanglai Gao and Xiangdong Su},
  booktitle = {Findings of the Association for Computational Linguistics: ACL 2026},
  year = {2026},
  doi = {10.18653/v1/2026.findings-acl.1072},
  url = {https://doi.org/10.18653/v1/2026.findings-acl.1072}
}

References

26 references, 6 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

  1. Hui Bai, Jan G. Voelkel, Shane Muldowney, Johannes C. Eichstaedt, and Robb Willer. 2025. Llm-generated messages can persuade humans on policy issues. Nature Communications, 16(1):1–12. find
  2. Simon Martin Breum, Daniel Vædele Egdal, Victor Gram Mortensen, Anders Giovanni Møller, and Luca Maria Aiello. 2024. The persuasive power of large language models. In Proceedings of the International AAAI Conference on Web and Social Media, volume 18, pages 152–163. find
  3. Chunkit Chan, Cheng Jiayang, Yauwai Yim, Zheye Deng, Wei Fan, Haoran Li, Xin Liu, Hongming Zhang, Weiqi Wang, and Yangqiu Song. 2024. NegotiationToM: A benchmark for stress-testing machine theory of mind on negotiation surrounding. In Findings of the Association for Computational Linguistics: EMNLP… find
  4. Guiming Hardy Chen, Shunian Chen, Ziche Liu, Feng Jiang, and Benyou Wang. 2024a. Humans or LLMs as the judge? a study on judgement bias. In Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing, pages 8301–8327, Miami, Florida, USA. Association for Computational Ling…
  5. Zhuang Chen, Jincenzi Wu, Jinfeng Zhou, Bosi Wen, Guanqun Bi, Gongyao Jiang, Yaru Cao, Mengting Hu, Yunghwei Lai, Zexuan Xiong, and Minlie Huang. 2024b. ToMBench: Benchmarking theory of mind in large language models. In Proceedings of the 62nd Annual Meeting of the Association for Computational Ling…
  6. Minje Choi, Jiaxin Pei, Sagar Kumar, Chang Shu, and David Jurgens. 2023. Do LLMs understand social knowledge? evaluating the sociability of large language models with SocKET benchmark. In Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing, pages 11370–11403, Singa… find
  7. Abhimanyu Dubey, Abhinav Jauhri, Abhinav Pandey, Abhishek Kadian, Ahmad Al-Dahle, Aiesha Letman, Akhil Mathur, Alan Schelten, Amy Yang, Angela Fan, Anirudh Goyal, Anthony S. Hartshorn, Aobo Yang, Archi Mitra, Archie Sravankumar, Artem Korenev, Arthur Hinsvark, Arun Rao, Aston Zhang, and 510 others.… — The Llama 3 Herd of Models
  8. Hyunwoo Kim, Melanie Sclar, Xuhui Zhou, Ronan Bras, Gunhee Kim, Yejin Choi, and Maarten Sap. 2023. FANToM: A benchmark for stress-testing machine theory of mind in interactions. In Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing, pages 14397–14413, Singapore. A… find
  9. Michal Kosinski. 2024. Evaluating large language models in theory of mind tasks. Proceedings of the National Academy of Sciences, 121(45):e2405460121.
  10. Tian Lan, Xiangdong Su, Xu Liu, Ruirui Wang, Ke Chang, Jiang Li, and Guanglai Gao. 2025. McBE: A multi-task Chinese bias evaluation benchmark for large language models. In Findings of the Association for Computational Linguistics: ACL 2025, pages 6033–6056, Vienna, Austria. Association for Computati… find
  11. Yang Liu, Dan Iter, Yichong Xu, Shuohang Wang, Ruochen Xu, and Chenguang Zhu. 2023. G-eval: NLG evaluation using gpt-4 with better human alignment. In Proceedings of the 2023 Conference on Empirical Methods in Natural Language Processing, pages 2511–2522, Singapore. Association for Computational Lin…
  12. Richard E. Petty and John T. Cacioppo. 1986. The Elaboration Likelihood Model of Persuasion, pages 1–24. Springer New York, New York, NY. find
  13. David Premack and Guy Woodruff. 1978. Does the chimpanzee have a theory of mind? Behavioral and Brain Sciences, 1(4):515–526. find
  14. Paula Rescala, Manoel Horta Ribeiro, Tiancheng Hu, and Robert West. 2024. Can language models recognize convincing arguments? In Findings of the Association for Computational Linguistics: EMNLP 2024, pages 8826–8837, Miami, Florida, USA. Association for Computational Linguistics. find
  15. Matthew Riemer, Zahra Ashktorab, Djallel Bouneffouf, Payel Das, Miao Liu, Justin D. Weisz, and Murray Campbell. 2024. Position: Theory of mind benchmarks are broken for large language models. find
  16. Francesco Salvi, Manoel Horta Ribeiro, Riccardo Gallotti, and Robert West. 2025. On the conversational persuasiveness of gpt-4. Nature Human Behaviour, 9(8):1645–1653. find
  17. Aman Singh Thakur, Kartik Choudhary, Venkat Srinik Ramayapally, Sankaran Vaidyanathan, and Dieuwke Hupkes. 2025. Judging the judges: Evaluating alignment and vulnerabilities in LLMs-as-judges. In Proceedings of the Fourth Workshop on Generation, Evaluation and Metrics (GEM²), pages 404–430, Vienna,… find
  18. Alexander Wan, Eric Wallace, and Dan Klein. 2024. What evidence do language models find convincing? In Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pages 7468–7484, Bangkok, Thailand. Association for Computational Linguistics. find
  19. Peiyi Wang, Lei Li, Liang Chen, Zefan Cai, Dawei Zhu, Binghuai Lin, Yunbo Cao, Lingpeng Kong, Qi Liu, Tianyu Liu, and Zhifang Sui. 2024. Large language models are not fair evaluators. In Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers),… find
  20. Hainiu Xu, Runcong Zhao, Lixing Zhu, Jinhua Du, and Yulan He. 2024. OpenToM: A comprehensive benchmark for evaluating theory-of-mind reasoning capabilities of large language models. In Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), pa… find
  21. Qwen An Yang, Baosong Yang, Beichen Zhang, Binyuan Hui, Bo Zheng, Bowen Yu, Chengyuan Li, Dayiheng Liu, Fei Huang, Guanting Dong, Haoran Wei, Huan Lin, Jian Yang, Jianhong Tu, Jianwei Zhang, Jianxin Yang, Jiaxin Yang, Jingren Zhou, Junyang Lin, and 25 others. 2024. Qwen2.5 technical report. ArXiv, a… find
  22. Fangxu Yu, Lai Jiang, Shenyi Huang, Zhen Wu, and Xinyu Dai. 2025. Persuasivetom: A benchmark for evaluating machine theory of mind in persuasive dialogues. ArXiv, abs/2502.21017. — [arXiv:2502.21017](https://arxiv.org/abs/2502.21017) find
  23. Jiarui Zhang. 2024. Guided profile generation improves personalization with large language models. In Findings of the Association for Computational Linguistics: EMNLP 2024, pages 4005–4016, Miami, Florida, USA. Association for Computational Linguistics. find
  24. Lianmin Zheng, Wei-Lin Chiang, Ying Sheng, Siyuan Zhuang, Zhanghao Wu, Yonghao Zhuang, Zi Lin, Zhuohan Li, Dacheng Li, Eric Xing, Hao Zhang, Joseph E Gonzalez, and Ion Stoica. 2023. Judging llm-as-a-judge with mt-bench and chatbot arena. In Advances in Neural Information Processing Systems, volume 3…
  25. Mingqian Zheng, Jiaxin Pei, Lajanugen Logeswaran, Moontae Lee, and David Jurgens. 2024. When ‘a helpful assistant” is not really helpful: Personas in system prompts do not improve performances of large language models. In Findings of the Association for Computational Linguistics: EMNLP 2024, pages 1… find
  26. Nana Zhu, Zixian Feng, Hang Wang, Xing Gao, Xinyi Wang, and Yuanxing Liu. 2025. Clarifying user preference with maximum entropy based recommendation. DATA INTELLIGENCE, 7(2):527–548. find

Updated

↑ back to top

This page in Markdown → https://md.research.elamur.ai/free/en/papers/wang-2026-tomelp-theory-mind-benchmark-route/