---
title: "Evaluating Answer Leakage Robustness of LLM Tutors against"
description: "The authors build a framework of multi-turn dialogues between a tutor, which knows the solution but is told not to reveal it, and… 41 references."
authors: ["Jin Zhao", "Marta Knežević", "Tanja Käser"]
year: 2026
venue: "Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)"
source: "https://aclanthology.org/2026.acl-long.1412/"
doi: "10.18653/v1/2026.acl-long.1412"
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/zhao-2026-evaluating-answer-leakage-robustness-llm/"
---
[Elamur Research](https://md.research.elamur.ai/free/en/) · [this page in HTML](https://research.elamur.ai/free/en/papers/zhao-2026-evaluating-answer-leakage-robustness-llm/)

[home](https://md.research.elamur.ai/free/en/) / [papers](https://md.research.elamur.ai/free/en/papers/) / [2026](https://md.research.elamur.ai/free/en/papers/2026/page/12/) / Evaluating Answer Leakage Robustness of LLM Tutors against Adversarial Student Attacks

# Evaluating Answer Leakage Robustness of LLM Tutors against Adversarial Student Attacks

**Authors:** [Jin Zhao](https://md.research.elamur.ai/free/en/authors/jin-zhao/), Marta Knežević and Tanja Käser · **Year:** 2026 · **Venue:** [Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)](https://md.research.elamur.ai/free/en/venues/acl/) ([ACL 2026](https://md.research.elamur.ai/free/en/venues/acl/2026/)) · **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/2026.acl-long.1412/) · [PDF](https://aclanthology.org/2026.acl-long.1412.pdf) · [DOI 10.18653/v1/2026.acl-long.1412](https://doi.org/10.18653/v1/2026.acl-long.1412) · [OpenAlex](https://openalex.org/W7166850939)

**Citations:** 1 citations (OpenAlex, 2026-10-14)

## Summary
LLM tutors are meant to scaffold learning rather than hand out solutions, but earlier evaluations of answer leakage mostly assumed cooperative learners. This paper asks how well such tutors resist students who deliberately try to extract the final answer. The authors build a framework of multi-turn dialogues between a tutor, which knows the solution but is told not to reveal it, and an adversarial student. LLM judges flag leakage turn by turn. They adapt six families of adversarial and persuasive techniques to tutoring and release a dataset of attack prompts. The evaluated tutors span a range of LLM families, models specifically tuned for pedagogy, and a setup built from multiple cooperating agents. Simple in-context student agents turn out to be weak attackers, since they often solve the problem themselves or start cooperating with the tutor. The authors therefore fine-tune a dedicated adversarial student agent and propose it as the basis of a standardized robustness benchmark. Persuasion-based strategies are the most effective attacks. Two lightweight tutor-side defenses, explicit reasoning and multi-agent refinement, reduce leakage substantially. This matters for deploying LLMs in homework and assessment settings.

## Key points
- The evaluation covers seven open-source LLM tutors and one closed-source model. It also includes the pedagogically aligned MathDial-SFT, SocraticLM and TutorRL-7B.
- Against the fine-tuned adversarial student, the aligned tutors leaked answers in 70%, 70% and 83% of dialogues, respectively. The student agent itself revealed the answer in only 1–2% of cases.
- Under the fine-tuned attacker, Llama-8B tutor leakage dropped from 34% to 6% with the multi-agent defense and to 3% with the reasoning defense. Qwen-32B fell from 70% to 4% and 3%. Qwen-7B improved less, from 82% to 61% and 38%.
- All 36 comparisons between defended tutor variants and the baseline in-context tutor were statistically significant.
- The framework generalizes beyond math. Reasoning tutors cut leakage from 88% to 10% on MCQ tasks and from 88% to 41% on coding tasks.

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

📚 Shelves (3): [🖥️ Computer Science](https://md.research.elamur.ai/free/en/shelves/computer-science/) · 16,667 works · [🤖 Artificial Intelligence](https://md.research.elamur.ai/free/en/shelves/computer-science/artificial-intelligence/) · 16,194 works · [🗄️ Information Systems](https://md.research.elamur.ai/free/en/shelves/computer-science/information-systems/) · 1,437 works

🧭 Topics (2): [🧠 Language models & pretraining](https://md.research.elamur.ai/free/en/topics/language-models/) · 3,320 works · [📏 Evaluation & benchmarks](https://md.research.elamur.ai/free/en/topics/evaluation-benchmarks/) · 1,220 works

## Cite

Cite as: Jin Zhao, Marta Knežević and Tanja Käser. 2026. Evaluating Answer Leakage Robustness of LLM Tutors against Adversarial Student Attacks. Proceedings of the 64th 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{zhao2026evaluating,
  title = {{Evaluating Answer Leakage Robustness of LLM Tutors against Adversarial Student Attacks}},
  author = {Jin Zhao, Marta Knežević and Tanja Käser},
  booktitle = {Proceedings of the 64th Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers)},
  year = {2026},
  doi = {10.18653/v1/2026.acl-long.1412},
  url = {https://doi.org/10.18653/v1/2026.acl-long.1412}
}
```

- [← previous Evaluating Adaptive Personalization of Educational Readings with Simulated Learners](https://md.research.elamur.ai/free/en/papers/woo-2026-evaluating-adaptive-personalization-educational-readings/)
- [next → Evaluating Document-Tuned Transformer Representations for Person-level Mental Health Assessment](https://md.research.elamur.ai/free/en/papers/marker-2026-evaluating-document-tuned-transformer-representations/)
- [↑ back to papers of 2026](https://md.research.elamur.ai/free/en/papers/2026/page/12/)

## References

41 references, 12 of them with a card of ours. ↗ leads to the work, «find» searches for one we have no id for yet.

1. Alon Albalak, Duy Phung, Nathan Lile, Rafael Rafailov, Kanishk Gandhi, Louis Castricato, Anikait Singh, Chase Blagden, Violet Xiang, Dakota Mahan, and 1 others. 2025. Big-math: A large-scale, high-quality math dataset for reinforcement learning in language models. arXiv preprint arXiv:2502.17387. [find](https://scholar.google.com/scholar?q=Alon+Albalak%2C+Duy+Phung%2C+Nathan+Lile%2C+Rafael+Rafailov%2C+Kanishk+Gandhi%2C+Louis+Castricato%2C+Anikait+Singh%2C+Chase+Blagden%2C+Violet+Xiang%2C+Dakota+Mahan%2C+and+1+others%2E+2025%2E+Big%2Dmath%3A+A+large%2Dscale%2C+high%2Dqua)
2. Maksym Andriushchenko, Alexandra Souly, Mateusz Dziemian, Derek Duenas, Maxwell Lin, Justin Wang, Dan Hendrycks, Andy Zou, J Zico Kolter, Matt Fredrikson, Yarin Gal, and Xander Davies. 2025. Agentharm: A benchmark for measuring harmfulness of LLM agents. In The Thirteenth International Conference on… [find](https://scholar.google.com/scholar?q=Maksym+Andriushchenko%2C+Alexandra+Souly%2C+Mateusz+Dziemian%2C+Derek+Duenas%2C+Maxwell+Lin%2C+Justin+Wang%2C+Dan+Hendrycks%2C+Andy+Zou%2C+J+Zico+Kolter%2C+Matt+Fredrikson%2C+Yarin+Gal%2C+and+Xander+Davies%2E+2025%2E+Agentharm)
3. Elizabeth L Bjork and Robert A Bjork. 2011. Making things hard on yourself, but in a good way: Creating desirable difficulties to enhance learning. Psychology and the real world: Essays illustrating fundamental contributions to society, 2(59-68):56–64. [find](https://scholar.google.com/scholar?q=Elizabeth+L+Bjork+and+Robert+A+Bjork%2E+2011%2E+Making+things+hard+on+yourself%2C+but+in+a+good+way%3A+Creating+desirable+difficulties+to+enhance+learning%2E+Psychology+and+the+real+world%3A+Essays+illustrating+f)
4. Robert A Bjork and Elizabeth L Bjork. 2020. Desirable difficulties in theory and practice. Journal of Applied research in Memory and Cognition, 9(4):475. [find](https://scholar.google.com/scholar?q=Robert+A+Bjork+and+Elizabeth+L+Bjork%2E+2020%2E+Desirable+difficulties+in+theory+and+practice%2E+Journal+of+Applied+research+in+Memory+and+Cognition%2C+9%284%29%3A475%2E)
5. Conrad Borchers, Cindy Peng, Qianru Lyu, Paulo F Carvalho, Kenneth R Koedinger, and Vincent Aleven. 2025. Student perceptions of adaptive goal setting recommendations: a design prototyping study. In International Conference on Artificial Intelligence in Education, pages 244–251. Springer. [find](https://scholar.google.com/scholar?q=Conrad+Borchers%2C+Cindy+Peng%2C+Qianru+Lyu%2C+Paulo+F+Carvalho%2C+Kenneth+R+Koedinger%2C+and+Vincent+Aleven%2E+2025%2E+Student+perceptions+of+adaptive+goal+setting+recommendations%3A+a+design+prototyping+study%2E+In+I)
6. Patrick Chao, Alexander Robey, Edgar Dobriban, Hamed Hassani, George J. Pappas, and Eric Wong. 2025. [Jailbreaking black box large language models in twenty queries](https://md.research.elamur.ai/free/en/papers/chao-2023-jailbreaking-black-box-large-language/). In 2025 IEEE Conference on Secure and Trustworthy Machine Learning (SaTML), pages 23–42.
7. Mark Chen, Jerry Tworek, Heewoo Jun, Qiming Yuan, Henrique Ponde de Oliveira Pinto, Jared Kaplan, Harri Edwards, Yuri Burda, Nicholas Joseph, Greg Brockman, Alex Ray, Raul Puri, Gretchen Krueger, Michael Petrov, Heidy Khlaaf, Girish Sastry, Pamela Mishkin, Brooke Chan, Scott Gray, and 39 others. 202… — [Evaluating Large Language Models Trained on Code](https://md.research.elamur.ai/free/en/papers/chen-2021-evaluating-large-language-models-trained/)
8. Yulin Chen, Ning Ding, Hai-Tao Zheng, Zhiyuan Liu, Maosong Sun, and Bowen Zhou. 2024. Empowering private tutoring by chaining large language models. In Proceedings of the 33rd ACM International Conference on Information and Knowledge Management, CIKM ’24, page 354–364, New York, NY, USA. Association… [find](https://scholar.google.com/scholar?q=Yulin+Chen%2C+Ning+Ding%2C+Hai%2DTao+Zheng%2C+Zhiyuan+Liu%2C+Maosong+Sun%2C+and+Bowen+Zhou%2E+2024%2E+Empowering+private+tutoring+by+chaining+large+language+models%2E+In+Proceedings+of+the+33rd+ACM+International+Confer)
9. Karl Cobbe, Vineet Kosaraju, Mohammad Bavarian, Mark Chen, Heewoo Jun, Lukasz Kaiser, Matthias Plappert, Jerry Tworek, Jacob Hilton, Reiichiro Nakano, and 1 others. 2021. [Training verifiers to solve math word problems](https://md.research.elamur.ai/free/en/papers/cobbe-2021-training-verifiers-solve-math-word/). arXiv preprint arXiv:2110.14168.
10. Nico Daheim, Jakub Macina, Manu Kapur, Iryna Gurevych, and Mrinmaya Sachan. 2024. Stepwise verification and remediation of student reasoning errors with large language model tutors. In Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing, pages 8386–8411, Miami, Flo… [find](https://scholar.google.com/scholar?q=Nico+Daheim%2C+Jakub+Macina%2C+Manu+Kapur%2C+Iryna+Gurevych%2C+and+Mrinmaya+Sachan%2E+2024%2E+Stepwise+verification+and+remediation+of+student+reasoning+errors+with+large+language+model+tutors%2E+In+Proceedings+of+)
11. David Dinucu-Jianu, Jakub Macina, Nico Daheim, Ido Hakimi, Iryna Gurevych, and Mrinmaya Sachan. 2025. From problem-solving to teaching problem-solving: Aligning LLMs with pedagogy using reinforcement learning. In Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing,… [find](https://scholar.google.com/scholar?q=David+Dinucu%2DJianu%2C+Jakub+Macina%2C+Nico+Daheim%2C+Ido+Hakimi%2C+Iryna+Gurevych%2C+and+Mrinmaya+Sachan%2E+2025%2E+From+problem%2Dsolving+to+teaching+problem%2Dsolving%3A+Aligning+LLMs+with+pedagogy+using+reinforcement+)
12. Fares Fawzi, Vinitra Swamy, Dominik Glandorf, Tanya Nazaretsky, and Tanja Käser. 2025. Scribe: Structured chain reasoning for interactive behaviour explanations using tool calling. In Proceedings of the 2025 Conference on Empirical Methods in Natural Language Processing, pages 29285–29310. [find](https://scholar.google.com/scholar?q=Fares+Fawzi%2C+Vinitra+Swamy%2C+Dominik+Glandorf%2C+Tanya+Nazaretsky%2C+and+Tanja+K%C3%A4ser%2E+2025%2E+Scribe%3A+Structured+chain+reasoning+for+interactive+behaviour+explanations+using+tool+calling%2E+In+Proceedings+of+t)
13. Aaron Grattafiori, Abhimanyu Dubey, Abhinav Jauhri, Abhinav Pandey, Abhishek Kadian, Ahmad Al-Dahle, Aiesha Letman, Akhil Mathur, Alan Schelten, Alex Vaughan, and 1 others. 2024. [The llama 3 herd of models](https://md.research.elamur.ai/free/en/papers/grattafiori-2024-llama-3-herd-models/). arXiv preprint arXiv:2407.21783.
14. Adit Gupta, Jennifer Reddig, Tommaso Calo, Daniel Weitekamp, and Christopher J MacLellan. 2025. Beyond final answers: Evaluating large language models for math tutoring. In International Conference on Artificial Intelligence in Education, pages 323–337. Springer. [find](https://scholar.google.com/scholar?q=Adit+Gupta%2C+Jennifer+Reddig%2C+Tommaso+Calo%2C+Daniel+Weitekamp%2C+and+Christopher+J+MacLellan%2E+2025%2E+Beyond+final+answers%3A+Evaluating+large+language+models+for+math+tutoring%2E+In+International+Conference+on)
15. Dan Hendrycks, Collin Burns, Steven Basart, Andy Zou, Mantas Mazeika, Dawn Song, and Jacob Steinhardt. 2021. [Measuring massive multitask language understanding](https://md.research.elamur.ai/free/en/papers/hendrycks-2020-measuring-massive-multitask-language-understanding/). In International Conference on Learning Representations.
16. Yilin Jiang, Mingzi Zhang, Xuanyu Yin, Sheng Jin, Suyu Lu, Zuocan Ying, Zengyi Yu, and Xiangjie Kong. 2026. Eduguardbench: A holistic benchmark for evaluating the pedagogical fidelity and adversarial safety of llms as simulated teachers. Proceedings of the AAAI Conference on Artificial Intelligence,… [find](https://scholar.google.com/scholar?q=Yilin+Jiang%2C+Mingzi+Zhang%2C+Xuanyu+Yin%2C+Sheng+Jin%2C+Suyu+Lu%2C+Zuocan+Ying%2C+Zengyi+Yu%2C+and+Xiangjie+Kong%2E+2026%2E+Eduguardbench%3A+A+holistic+benchmark+for+evaluating+the+pedagogical+fidelity+and+adversarial+)
17. Irina Jurenka, Markus Kunesch, Kevin R McKee, Daniel Gillick, Shaojian Zhu, Sara Wiltberger, Shubham Milind Phal, Katherine Hermann, Daniel Kasenberg, Avishkar Bhoopchand, and 1 others. 2024. Towards responsible development of generative ai for education: An evaluation-driven approach. arXiv preprin… [find](https://scholar.google.com/scholar?q=Irina+Jurenka%2C+Markus+Kunesch%2C+Kevin+R+McKee%2C+Daniel+Gillick%2C+Shaojian+Zhu%2C+Sara+Wiltberger%2C+Shubham+Milind+Phal%2C+Katherine+Hermann%2C+Daniel+Kasenberg%2C+Avishkar+Bhoopchand%2C+and+1+others%2E+2024%2E+Towards+)
18. Woosuk Kwon, Zhuohan Li, Siyuan Zhuang, Ying Sheng, Lianmin Zheng, Cody Hao Yu, Joseph Gonzalez, Hao Zhang, and Ion Stoica. 2023. [Efficient memory management for large language model serving with pagedattention](https://md.research.elamur.ai/free/en/papers/kwon-2023-efficient-memory-management-large-language/). In Proceedings of the 29th Symposium on Operating Systems Principles, SOSP ’23, page 611…
19. Jiayu Liu, Zhenya Huang, Tong Xiao, Jing Sha, Jinze Wu, Qi Liu, Shijin Wang, and Enhong Chen. 2024. Socraticlm: Exploring socratic personalized teaching with large language models. Advances in Neural Information Processing Systems, 37:85693–85721. [find](https://scholar.google.com/scholar?q=Jiayu+Liu%2C+Zhenya+Huang%2C+Tong+Xiao%2C+Jing+Sha%2C+Jinze+Wu%2C+Qi+Liu%2C+Shijin+Wang%2C+and+Enhong+Chen%2E+2024%2E+Socraticlm%3A+Exploring+socratic+personalized+teaching+with+large+language+models%2E+Advances+in+Neural+)
20. Jakub Macina, Nico Daheim, Sankalan Chowdhury, Tanmay Sinha, Manu Kapur, Iryna Gurevych, and Mrinmaya Sachan. 2023. [MathDial: A dialogue tutoring dataset with rich pedagogical properties grounded in math reasoning problems](https://md.research.elamur.ai/free/en/papers/macina-2023-mathdial-dialogue-tutoring-dataset-rich/). In Findings of the Association for Computational Linguistics: EMNLP 2023, pa…
21. Kaushal Kumar Maurya, Kv Aditya Srivatsa, Kseniia Petukhova, and Ekaterina Kochmar. 2025. Unifying AI tutor evaluation: An evaluation taxonomy for pedagogical ability assessment of LLM-powered AI tutors. In Proceedings of the 2025 Conference of the Nations of the Americas Chapter of the Association… [find](https://scholar.google.com/scholar?q=Kaushal+Kumar+Maurya%2C+Kv+Aditya+Srivatsa%2C+Kseniia+Petukhova%2C+and+Ekaterina+Kochmar%2E+2025%2E+Unifying+AI+tutor+evaluation%3A+An+evaluation+taxonomy+for+pedagogical+ability+assessment+of+LLM%2Dpowered+AI+tuto)
22. Qurat-ul-Ann Mirza and N Sanjay Rebello. 2025. Help or hype? students’ engagement and perception of using ai to solve physics problems. arXiv preprint arXiv:2509.25642. — [arXiv:2509.25642](https://arxiv.org/abs/2509.25642) [find](https://scholar.google.com/scholar?q=Qurat%2Dul%2DAnn+Mirza+and+N+Sanjay+Rebello%2E+2025%2E+Help+or+hype%3F+students%E2%80%99+engagement+and+perception+of+using+ai+to+solve+physics+problems%2E+arXiv+preprint+arXiv%3A2509%2E25642%2E+%E2%80%94+%5BarXiv%3A2509%2E25642%5D%28https%3A%2F%2Far)
23. Tanya Nazaretsky, Paola Mejia-Domenzain, Vinitra Swamy, Jibril Frej, and Tanja Käser. 2024. Ai or human? evaluating student feedback perceptions in higher education. In European conference on technology enhanced learning, pages 284–298. Springer. [find](https://scholar.google.com/scholar?q=Tanya+Nazaretsky%2C+Paola+Mejia%2DDomenzain%2C+Vinitra+Swamy%2C+Jibril+Frej%2C+and+Tanja+K%C3%A4ser%2E+2024%2E+Ai+or+human%3F+evaluating+student+feedback+perceptions+in+higher+education%2E+In+European+conference+on+technolo)
24. Zachary A Pardos and Shreya Bhandari. 2023. Learning gain differences between chatgpt and human tutor generated algebra hints. arXiv preprint arXiv:2302.06871. [find](https://scholar.google.com/scholar?q=Zachary+A+Pardos+and+Shreya+Bhandari%2E+2023%2E+Learning+gain+differences+between+chatgpt+and+human+tutor+generated+algebra+hints%2E+arXiv+preprint+arXiv%3A2302%2E06871%2E)
25. Minju Park, Sojung Kim, Seunghyun Lee, Soonwoo Kwon, and Kyuseok Kim. 2024. Empowering personalized learning through a conversation-based tutoring system with student modeling. In Extended Abstracts of the CHI Conference on Human Factors in Computing Systems, CHI ’24, page 1–10. ACM. [find](https://scholar.google.com/scholar?q=Minju+Park%2C+Sojung+Kim%2C+Seunghyun+Lee%2C+Soonwoo+Kwon%2C+and+Kyuseok+Kim%2E+2024%2E+Empowering+personalized+learning+through+a+conversation%2Dbased+tutoring+system+with+student+modeling%2E+In+Extended+Abstracts+o)
26. Janos Perczel, Jin Chow, and Dorottya Demszky. 2025. Teachlm: Post-training llms for education using authentic learning data. Preprint, arXiv:2510.05087. — [arXiv:2510.05087](https://arxiv.org/abs/2510.05087) [find](https://scholar.google.com/scholar?q=Janos+Perczel%2C+Jin+Chow%2C+and+Dorottya+Demszky%2E+2025%2E+Teachlm%3A+Post%2Dtraining+llms+for+education+using+authentic+learning+data%2E+Preprint%2C+arXiv%3A2510%2E05087%2E+%E2%80%94+%5BarXiv%3A2510%2E05087%5D%28https%3A%2F%2Farxiv%2Eorg%2Fabs%2F251)
27. Romain Puech, Jakub Macina, Julia Chatain, Mrinmaya Sachan, and Manu Kapur. 2025. Towards the pedagogical steering of large language models for tutoring: A case study with modeling productive failure. In Findings of the Association for Computational Linguistics: ACL 2025, pages 26291–26311, Vienna,… [find](https://scholar.google.com/scholar?q=Romain+Puech%2C+Jakub+Macina%2C+Julia+Chatain%2C+Mrinmaya+Sachan%2C+and+Manu+Kapur%2E+2025%2E+Towards+the+pedagogical+steering+of+large+language+models+for+tutoring%3A+A+case+study+with+modeling+productive+failure%2E)
28. Qibing Ren, Hao Li, Dongrui Liu, Zhanxu Xie, Xiaoya Lu, Yu Qiao, Lei Sha, Junchi Yan, Lizhuang Ma, and Jing Shao. 2025. [Llms know their vulnerabilities: Uncover safety gaps through natural distribution shifts](https://md.research.elamur.ai/free/en/papers/ren-2025-llms-know-their-vulnerabilities-uncover/). In Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics…
29. Lianne Roest, Hieke Keuning, and Johan Jeuring. 2024. Next-step hint generation for introductory programming using large language models. In Proceedings of the 26th Australasian Computing Education Conference, pages 144–153. [find](https://scholar.google.com/scholar?q=Lianne+Roest%2C+Hieke+Keuning%2C+and+Johan+Jeuring%2E+2024%2E+Next%2Dstep+hint+generation+for+introductory+programming+using+large+language+models%2E+In+Proceedings+of+the+26th+Australasian+Computing+Education+Co)
30. Alexander Scarlatos, Naiming Liu, Jaewook Lee, Richard Baraniuk, and Andrew Lan. 2025. Training llm-based tutors to improve student learning outcomes in dialogues. In Artificial Intelligence in Education: 26th International Conference, AIED 2025, Palermo, Italy, July 22–26, 2025, Proceedings, Part I… [find](https://scholar.google.com/scholar?q=Alexander+Scarlatos%2C+Naiming+Liu%2C+Jaewook+Lee%2C+Richard+Baraniuk%2C+and+Andrew+Lan%2E+2025%2E+Training+llm%2Dbased+tutors+to+improve+student+learning+outcomes+in+dialogues%2E+In+Artificial+Intelligence+in+Educat)
31. Yao Shi, Rongkeng Liang, and Yong Xu. 2025. [Educationq: Evaluating llms’ teaching capabilities through multi-agent dialogue framework](https://md.research.elamur.ai/free/en/papers/shi-2025-educationq-evaluating-llms-teaching-capabilities/). In Proceedings of the 63rd Annual Meeting of the Association for Computational Linguistics (Volume 1: Long Papers), page 32799–32828. Association for Computational L…
32. Shashank Sonkar, Naiming Liu, Debshila Mallick, and Richard Baraniuk. 2023. CLASS: A design framework for building intelligent tutoring systems based on learning science principles. In Findings of the Association for Computational Linguistics: EMNLP 2023, pages 1941–1961, Singapore. Association for… [find](https://scholar.google.com/scholar?q=Shashank+Sonkar%2C+Naiming+Liu%2C+Debshila+Mallick%2C+and+Richard+Baraniuk%2E+2023%2E+CLASS%3A+A+design+framework+for+building+intelligent+tutoring+systems+based+on+learning+science+principles%2E+In+Findings+of+the)
33. Shashank Sonkar, Kangqi Ni, Sapana Chaudhary, and Richard Baraniuk. 2024. Pedagogical alignment of large language models. In Findings of the Association for Computational Linguistics: EMNLP 2024, pages 13641–13650, Miami, Florida, USA. Association for Computational Linguistics. [find](https://scholar.google.com/scholar?q=Shashank+Sonkar%2C+Kangqi+Ni%2C+Sapana+Chaudhary%2C+and+Richard+Baraniuk%2E+2024%2E+Pedagogical+alignment+of+large+language+models%2E+In+Findings+of+the+Association+for+Computational+Linguistics%3A+EMNLP+2024%2C+page)
34. Rakshith S Srinivasa, Zora Che, Chen Bo Calvin Zhang, Diego Mares, Ernesto Hernandez, Jayeon Park, Dean Lee, Guillermo Mangialardi, Charmaine Ng, Ed-Yeremai Hernandez Cardona, Anisha Gunjal, Yunzhong He, Bing Liu, and Chen Xing. 2025. Tutor-bench: A benchmark to assess tutoring capabilities of large… — [arXiv:2510.02663](https://arxiv.org/abs/2510.02663) [find](https://scholar.google.com/scholar?q=Rakshith+S+Srinivasa%2C+Zora+Che%2C+Chen+Bo+Calvin+Zhang%2C+Diego+Mares%2C+Ernesto+Hernandez%2C+Jayeon+Park%2C+Dean+Lee%2C+Guillermo+Mangialardi%2C+Charmaine+Ng%2C+Ed%2DYeremai+Hernandez+Cardona%2C+Anisha+Gunjal%2C+Yunzhong+)
35. Anaïs Tack and Chris Piech. 2022. The AI teacher test: Measuring the pedagogical ability of blender and GPT-3 in educational dialogues. In Proceedings of the 15th International Conference on Educational Data Mining, pages 522–529, Durham, United Kingdom. International Educational Data Mining Society… [find](https://scholar.google.com/scholar?q=Ana%C3%AFs+Tack+and+Chris+Piech%2E+2022%2E+The+AI+teacher+test%3A+Measuring+the+pedagogical+ability+of+blender+and+GPT%2D3+in+educational+dialogues%2E+In+Proceedings+of+the+15th+International+Conference+on+Education)
36. Rose Wang, Qingyang Zhang, Carly Robinson, Susanna Loeb, and Dorottya Demszky. 2024. Bridging the novice-expert gap via models of decision-making: A case study on remediating math mistakes. In Proceedings of the 2024 Conference of the North American Chapter of the Association for Computational Lingu… [find](https://scholar.google.com/scholar?q=Rose+Wang%2C+Qingyang+Zhang%2C+Carly+Robinson%2C+Susanna+Loeb%2C+and+Dorottya+Demszky%2E+2024%2E+Bridging+the+novice%2Dexpert+gap+via+models+of+decision%2Dmaking%3A+A+case+study+on+remediating+math+mistakes%2E+In+Proceed)
37. Jason Wei, Xuezhi Wang, Dale Schuurmans, Maarten Bosma, Fei Xia, Ed Chi, Quoc V Le, Denny Zhou, and 1 others. 2022. [Chain-of-thought prompting elicits reasoning in large language models](https://md.research.elamur.ai/free/en/papers/wei-2022-chain-thought-prompting-elicits-reasoning/). Advances in neural information processing systems, 35:24824– 24837.
38. Qwen: An Yang, Baosong Yang, Beichen Zhang, Binyuan Hui, Bo Zheng, Bowen Yu, Chengyuan Li, Dayiheng Liu, Fei Huang, Haoran Wei, Huan Lin, Jian Yang, Jianhong Tu, Jianwei Zhang, Jianxin Yang, Jiaxi Yang, Jingren Zhou, Junyang Lin, Kai Dang, and 23 others. 2025. Qwen2.5 technical report. Preprint, arX… [find](https://scholar.google.com/scholar?q=Qwen%3A+An+Yang%2C+Baosong+Yang%2C+Beichen+Zhang%2C+Binyuan+Hui%2C+Bo+Zheng%2C+Bowen+Yu%2C+Chengyuan+Li%2C+Dayiheng+Liu%2C+Fei+Huang%2C+Haoran+Wei%2C+Huan+Lin%2C+Jian+Yang%2C+Jianhong+Tu%2C+Jianwei+Zhang%2C+Jianxin+Yang%2C+Jiaxi+Yan)
39. Shuzhou Yuan, William LaCroix, Hardik Ghoshal, Ercong Nie, and Michael Färber. 2025. Co-dae: Adapting large language models for education via chain-of-thought data augmentation. Preprint, arXiv:2508.08386. [find](https://scholar.google.com/scholar?q=Shuzhou+Yuan%2C+William+LaCroix%2C+Hardik+Ghoshal%2C+Ercong+Nie%2C+and+Michael+F%C3%A4rber%2E+2025%2E+Co%2Ddae%3A+Adapting+large+language+models+for+education+via+chain%2Dof%2Dthought+data+augmentation%2E+Preprint%2C+arXiv%3A2508%2E0)
40. Yi Zeng, Hongpeng Lin, Jingwen Zhang, Diyi Yang, Ruoxi Jia, and Weiyan Shi. 2024. [How johnny can persuade LLMs to jailbreak them: Rethinking persuasion to challenge AI safety by humanizing LLMs](https://md.research.elamur.ai/free/en/papers/zeng-2024-how-johnny-can-persuade-llms/). In Proceedings of the 62nd Annual Meeting of the Association for Computational Linguistics (Volume 1: Lon…
41. 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](https://md.research.elamur.ai/free/en/papers/zheng-2023-judging-llm-judge-mt-bench/). In Thirty-seventh Conference on Neural Information Processing…

*Updated 1 November 2026*