Empowering Vocabulary Learning Through Teaching AI: Using LLMs as a Student to Perform Learning by Teaching in Vocabulary Acquisition

Andreas Dengel; Andrew Vargo; Ayaka Sugawara; Koichi Kise; Ko Watanabe; Ralph L. Rose; Shoya Ishimaru; Tokio Uchida

arxiv: 2604.17893 · v1 · submitted 2026-04-20 · 💻 cs.HC

Empowering Vocabulary Learning Through Teaching AI: Using LLMs as a Student to Perform Learning by Teaching in Vocabulary Acquisition

Tokio Uchida , Ko Watanabe , Andrew Vargo , Shoya Ishimaru , Ralph L. Rose , Ayaka Sugawara , Andreas Dengel , Koichi Kise This is my paper

Pith reviewed 2026-05-10 04:26 UTC · model grok-4.3

classification 💻 cs.HC

keywords vocabulary learninglearning by teachinglarge language modelsAI in educationmemory retentioneducational technologyquestion generation

0 comments

The pith

Learners who teach vocabulary to an LLM student retain the words better at three and seven days than with standard study methods.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper tests a learning-by-teaching setup in which an LLM acts as the student and poses dynamic questions that prompt the human learner to explain vocabulary items. Ten participants who used this system showed stronger recall of the words after three and seven days compared with those who studied the same words through conventional methods. A sympathetic reader would care because the approach replaces rigid templates and human tutors with an on-demand AI that can generate contextually relevant questions on the fly. The work also notes that certain learner traits predict larger gains, suggesting the method could be adapted to individual needs.

Core claim

Participants who answered questions generated by an LLM configured as a student achieved better delayed recall of English vocabulary than participants who used traditional study methods. The LLM produces contextually relevant questions that help the learner identify gaps and reinforce knowledge while teaching the artificial student. The authors report measurable retention advantages at the three-day and seven-day tests and observe correlations between learner characteristics and the size of the benefit.

What carries the argument

An LLM acting as a student that generates dynamic, contextually relevant questions for the human learner to answer while teaching it vocabulary.

If this is right

Delayed recall of vocabulary items improves relative to conventional study.
Question generation for learning-by-teaching becomes feasible without hand-coded templates.
Learner traits can be used to identify who is likely to benefit most from the interaction.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

The same LLM-student format could be applied to factual material outside vocabulary, such as historical dates or scientific definitions.
Larger trials that equate total time on task would clarify whether the teaching step itself drives the gains.
Embedding the system in mobile apps could make learning-by-teaching available without requiring additional human partners.

Load-bearing premise

Any retention advantage comes from the teaching interaction with the AI rather than from extra practice time or the novelty of using new technology.

What would settle it

A follow-up experiment that gives both the AI-teaching group and a control group identical total study time, then measures whether the retention difference at three and seven days disappears.

Figures

Figures reproduced from arXiv: 2604.17893 by Andreas Dengel, Andrew Vargo, Ayaka Sugawara, Koichi Kise, Ko Watanabe, Ralph L. Rose, Shoya Ishimaru, Tokio Uchida.

**Figure 3.** Figure 3: Difference in the percentage of correct answers [PITH_FULL_IMAGE:figures/full_fig_p004_3.png] view at source ↗

**Figure 4.** Figure 4: Prompt feed to GPT-4o to generate material for [PITH_FULL_IMAGE:figures/full_fig_p006_4.png] view at source ↗

**Figure 6.** Figure 6: Baseline System: Learning without LLMs [PITH_FULL_IMAGE:figures/full_fig_p008_6.png] view at source ↗

**Figure 7.** Figure 7: Proposed System: Learning with LLMs [PITH_FULL_IMAGE:figures/full_fig_p008_7.png] view at source ↗

**Figure 9.** Figure 9: The average number of words entered per in [PITH_FULL_IMAGE:figures/full_fig_p009_9.png] view at source ↗

**Figure 8.** Figure 8: User interface of the pretest questions. Users [PITH_FULL_IMAGE:figures/full_fig_p009_8.png] view at source ↗

read the original abstract

"Learning by Teaching (LbT)" helps learners deepen their understanding by explaining concepts to others, with questions playing a vital role in identifying knowledge gaps and reinforcing comprehension. However, existing systems for generating such questions often rely on rigid templates and are expensive to build. To overcome these limitations, we developed a system using Large Language Models (LLMs) to create dynamic, contextually relevant questions for LbT. In our English vocabulary learning study, we examined which learner characteristics best leverage the system's benefits. Our results showed improved memory retention over traditional methods at three and seven days of testing, with ten participants. Additionally, we identified traits linked to better learning outcomes, highlighting the potential for tailored approaches. These findings support the development of scalable, cost-effective solutions to enhance LbT methods across various fields.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Desk Editor's Note private letter to a colleague

The paper uses LLMs to generate dynamic questions for learning-by-teaching in vocabulary tasks and reports retention gains in a 10-person study, but the small sample and missing controls make the main result hard to trust.

read the letter

The main thing to know is that this work replaces rigid templates with LLMs to produce context-aware questions inside a learning-by-teaching loop for English vocabulary, then runs a small user study claiming better retention at three and seven days than traditional methods. It also flags some learner traits that seem to predict stronger outcomes. That combination is new enough to be worth noticing, since earlier LbT systems were limited by how expensive and inflexible the question generators were. The paper does a reasonable job laying out the system architecture and showing that LLMs can handle the dynamic part without much hand-coding. It also tries to move beyond a simple yes/no result by looking at individual differences, which is a step in the right direction for making these tools practical. The soft spot is the evidence. Ten participants is a very thin base for claiming a retention advantage, and the description gives no clear account of how total time on task was equated, whether conditions were counterbalanced, what statistical tests were used, or whether anyone checked for simple novelty effects from talking to an AI. Without those pieces, the difference could easily come from extra engagement rather than the teaching mechanism itself. The abstract states the result but leaves the methods too thin to evaluate independently. This paper is mainly for people working on AI-supported language tools or HCI applications in education who want an early example of LLM-driven LbT. A reader hunting for ready-to-use systems or large-scale validation will not find it here. I would send it to peer review. The core idea is straightforward and the implementation looks workable, but referees can insist on a larger sample and tighter controls before the retention claim can be taken seriously.

Referee Report

2 major / 2 minor

Summary. The paper introduces an LLM-based system for dynamically generating questions to support Learning by Teaching (LbT) in English vocabulary acquisition, overcoming limitations of rigid template-based approaches. It reports results from a 10-participant study claiming superior long-term memory retention (at 3- and 7-day tests) relative to traditional methods, along with identification of learner traits that moderate benefits.

Significance. The core idea of leveraging LLMs for scalable, context-aware question generation in LbT is promising and could lower barriers to implementing effective teaching-as-learning strategies in HCI and education technology. The attention to individual learner characteristics is a constructive step toward personalization. However, the small sample and incomplete methodological transparency currently constrain the work's ability to influence practice or theory.

major comments (2)

[Abstract] The central retention claim (Abstract) rests on a 10-participant comparison whose methods, controls, statistical tests, and exclusion rules are not described. Without these details it is impossible to evaluate whether the reported advantage can be attributed to the LLM-LbT mechanism rather than confounds such as unequal time-on-task or AI novelty.
[Study / Methods section] No information is supplied on randomization or counterbalancing of conditions, baseline vocabulary pre-tests, total exposure time per condition, or any manipulation check for novelty effects. These omissions are load-bearing because the sole empirical support for the paper's contribution is the retention difference between conditions.

minor comments (2)

[Abstract] The abstract states that specific learner traits were linked to better outcomes but does not name them; adding this information would improve informativeness without lengthening the abstract substantially.
[Results section] A summary table of retention scores, participant demographics, and statistical results would greatly aid readability and allow readers to assess the magnitude of the reported effects.

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the constructive feedback and for recognizing the promise of using LLMs to support scalable Learning by Teaching. We agree that the original submission lacked sufficient methodological detail to allow readers to evaluate the retention findings. We have revised the manuscript to address these concerns directly.

read point-by-point responses

Referee: [Abstract] The central retention claim (Abstract) rests on a 10-participant comparison whose methods, controls, statistical tests, and exclusion rules are not described. Without these details it is impossible to evaluate whether the reported advantage can be attributed to the LLM-LbT mechanism rather than confounds such as unequal time-on-task or AI novelty.

Authors: We agree that the abstract and Methods section in the submitted version did not provide adequate information on the study procedures. In the revised manuscript we have expanded the Methods section to fully describe the experimental design, controls for time-on-task and novelty effects, the statistical tests performed on the 3-day and 7-day retention scores, and the rules applied for data exclusion. We have also updated the abstract to reference these methodological elements so that the retention claim can be properly assessed. revision: yes
Referee: [Study / Methods section] No information is supplied on randomization or counterbalancing of conditions, baseline vocabulary pre-tests, total exposure time per condition, or any manipulation check for novelty effects. These omissions are load-bearing because the sole empirical support for the paper's contribution is the retention difference between conditions.

Authors: We acknowledge the omission. The revised Methods section now supplies the missing information: details on how conditions were randomized and counterbalanced, the baseline vocabulary pre-tests that were administered, the recorded total exposure time per condition, and the post-experiment check used to assess novelty effects. These additions make the empirical comparison transparent and allow readers to judge whether the retention advantage can be attributed to the LLM-supported LbT mechanism. revision: yes

Circularity Check

0 steps flagged

No circularity: empirical user study with no derivation chain

full rationale

The paper reports results from a small-scale empirical user study (N=10) comparing LLM-generated questions in a Learning-by-Teaching condition against traditional vocabulary methods, measuring retention at 3- and 7-day intervals. No equations, first-principles derivations, fitted parameters, or mathematical predictions appear in the provided abstract or described content. The central claim rests on observed experimental outcomes rather than any reduction of results to inputs defined inside the paper. Self-citations, if present, are not load-bearing for any claimed uniqueness theorem or ansatz. This is a standard non-circular empirical report.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

Abstract-only review; no explicit free parameters, axioms, or invented entities are stated. The work relies on standard assumptions from educational psychology and LLM capabilities that are not audited here.

pith-pipeline@v0.9.0 · 5465 in / 1023 out tokens · 31339 ms · 2026-05-10T04:26:15.049989+00:00 · methodology

discussion (0)

Reference graph

Works this paper leans on

31 extracted references · 31 canonical work pages

[2]

Eleanor R Bowyer and Sebastian CK Shaw. 2021. Informal near-peer teaching in medical education: a scoping review.Education for Health34, 1 (2021), 29–33

work page 2021
[3]

HinMingFrankie Chik. 2021. Liji (Book of Rites). https://doi.org/10.14288/ 1.0404466

work page 2021
[4]

2014.The Book of Rites (Liji): Bilingual Edition, English and Chinese

Confucius (Attributed). 2014.The Book of Rites (Liji): Bilingual Edition, English and Chinese. James Legge. https://www.amazon.de/Book-Rites- Liji-Bilingual-English-ebook/dp/B00KVGYS9M Bilingual Edition, English and Chinese

work page 2014
[5]

Claudio G. Cortese. 2005. Learning through Teaching.Management Learn- ing36, 1 (2005), 87–115. https://doi.org/10.1177/1350507605049905

work page doi:10.1177/1350507605049905 2005
[6]

Amy Debbané, Ken Jen Lee, Jarvis Tse, and Edith Law. 2023. Learning by Teaching: Key Challenges and Design Implications.Proc. ACM Hum.- Comput. Interact.7, CSCW1 (April 2023), 1–34. https://doi.org/10.1145/ 3579501

work page 2023
[7]

Jiexin Ding, Bowen Zhao, Yuqi Huang, Yuntao Wang, and Yuanchun Shi

work page
[8]

InExtended Abstracts of the 2023 CHI Conference on Human Factors in Computing Systems(Hamburg, Germany) (CHI EA ’23)

GazeReader: Detecting Unknown Word Using Webcam for English as a Second Language (ESL) Learners. InExtended Abstracts of the 2023 CHI Conference on Human Factors in Computing Systems(Hamburg, Germany) (CHI EA ’23). Association for Computing Machinery, New York, NY, USA, Article 149, 7 pages. https://doi.org/10.1145/3544549.3585790

work page doi:10.1145/3544549.3585790 2023
[9]

Min Fan, Sheng Jin, and Alissa N. Antle. 2018. Designing Colours and Materials in Tangible Reading Products for Foreign Language Learners of English. InExtended Abstracts of the 2018 CHI Conference on Human Factors Empowering Vocabulary Learning Through Teaching AI: Using LLMs as a Student to Perform Learning by Teaching in Vocabulary Acquisition AHs 2026...

work page arXiv 2018
[10]

Logan Fiorella and Richard E. Mayer. 2013. The relative benefits of learning by teaching and teaching expectancy.Contemporary Educational Psychol- ogy38, 4 (2013), 281–288. https://doi.org/10.1016/j.cedpsych.2013.06.001

work page doi:10.1016/j.cedpsych.2013.06.001 2013
[11]

R. C. Gardner and P. D. MacIntyre. 1991. An Instrumental Motivation In Language Study: Who Says It Isn’t Effective?Studies in Second Language Acquisition13, 1 (1991), 57–72. https://doi.org/10.1017/S0272263100009724

work page doi:10.1017/s0272263100009724 1991
[12]

Taichi Higasa, Keitaro Tanaka, Qi Feng, and Shigeo Morishima. 2024. Keep Eyes on the Sentence: An Interactive Sentence Simplification System for English Learners Based on Eye Tracking and Large Language Models. In Extended Abstracts of the CHI Conference on Human Factors in Computing Systems(Honolulu, HI, USA)(CHI EA ’24). Association for Computing Machin...

work page arXiv 2024
[13]

Riku Higashimura, Ko Watanabe, Andrew Vargo, Motoi Iwata, Andreas Dengel, and Koichi Kise. 2024. Estimating Unknown English Words From User Smartphone Reading Behaviors.IEEE Access12 (2024), 140223–140234. https://doi.org/10.1109/ACCESS.2024.3457510

work page doi:10.1109/access.2024.3457510 2024
[14]

Hyoungwook Jin, Seonghee Lee, Hyungyu Shin, and Juho Kim. 2024. Teach AI How to Code: Using Large Language Models as Teachable Agents for Programming Education. InProceedings of the 2024 CHI Conference on Human Factors in Computing Systems (CHI ’24). Association for Computing Machinery, New York, NY, USA, 1—-28. https://doi.org/10.1145/3613904. 3642349

work page doi:10.1145/3613904 2024
[15]

Nayoung Jin and Hana Lee. 2022. StuBot: Learning by Teaching a Con- versational Agent Through Machine Reading Comprehension. InFind- ings of the Association for Computational Linguistics: EMNLP 2022, Yoav Goldberg, Zornitsa Kozareva, and Yue Zhang (Eds.). Association for Com- putational Linguistics, Abu Dhabi, United Arab Emirates, 3008–3020. https://doi....

work page doi:10.18653/v1/2022.findings-emnlp.219 2022
[16]

Pantasdo, Jessy Ceha, Sangho Suh, and Nicole Dillen

Edith Law, Parastoo Baghaei Ravari, Nalin Chhibber, Dana Kulic, Stephanie Lin, Kevin D. Pantasdo, Jessy Ceha, Sangho Suh, and Nicole Dillen. 2020. Curiosity Notebook: A Platform for Learning by Teaching Conversational Agents. InExtended Abstracts of the 2020 CHI Conference on Human Factors in Computing Systems (CHI EA ’20). Association for Computing Machi...

work page doi:10.1145/3334480.3382783 2020
[17]

Ziyi Liu, Zhengzhe Zhu, Lijun Zhu, Enze Jiang, Xiyun Hu, Kylie A Pep- pler, and Karthik Ramani. 2024. ClassMeta: Designing Interactive Vir- tual Classmate to Promote VR Classroom Participation. InProceedings of the 2024 CHI Conference on Human Factors in Computing Systems (CHI ’24). Association for Computing Machinery, New York, NY, USA, 1—-17. https://do...

work page doi:10.1145/3613904.3642947 2024
[18]

Ali Malik, Juliette Woodrow, and Chris Piech. 2024. Learners Teaching Novices: An Uplifting Alternative Assessment. InProceedings of the 55th ACM Technical Symposium on Computer Science Education V. 1 (SIGCSE 2024). Association for Computing Machinery, New York, NY, USA, 785—-

work page 2024
[19]

https://doi.org/10.1145/3626252.3630887

work page doi:10.1145/3626252.3630887
[20]

Markel, Steven G

Julia M. Markel, Steven G. Opferman, James A. Landay, and Chris Piech

work page
[21]

In Proceedings of the Tenth ACM Conference on Learning @ Scale (L@S ’23)

GPTeach: Interactive TA Training with GPT-based Students. In Proceedings of the Tenth ACM Conference on Learning @ Scale (L@S ’23). Association for Computing Machinery, New York, NY, USA, 226—-236. https://doi.org/10.1145/3573051.3593393

work page doi:10.1145/3573051.3593393
[22]

Noboru Matsuda. 2022. Teachable Agent as an Interactive Tool for Cog- nitive Task Analysis: A Case Study for Authoring an Expert Model.In- ternational Journal of Artificial Intelligence in Education32 (2022), 48–75. https://doi.org/10.1007/s40593-021-00265-z

work page doi:10.1007/s40593-021-00265-z 2022
[23]

Maximiliano Paredes-Velasco, Isaac Lozano-Osorio, Diana Pérez-Marín, and Liliana Patricia Santacruz-Valencia. 2024. A Case Study on Learn- ing Visual Programming With TutoApp for Composition of Tutorials: An Approach for Learning by Teaching.IEEE Transactions on Learning Technologies17 (2024), 498–513. https://doi.org/10.1109/TLT.2022.3226122

work page doi:10.1109/tlt.2022.3226122 2024
[24]

Nihar Sabnis and Tomohiro Nagashima. 2024. Empowering Learners: Chatbot-Mediated ’Learning-by-Teaching’. InExtended Abstracts of the CHI Conference on Human Factors in Computing Systems (CHI EA ’24). Association for Computing Machinery, New York, NY, USA, 1—-9. https: //doi.org/10.1145/3613905.3650754

work page doi:10.1145/3613905.3650754 2024
[25]

Fuxing Wang, Meixia Cheng, and Richard Mayer. 2023. Improving learning- by-teaching without audience interaction as a generative learning activity by minimizing the social presence of the audience.Journal of Educational Psychology115, 6 (2023), 783–797. https://doi.org/10.1037/edu0000801

work page doi:10.1037/edu0000801 2023
[26]

Ko Watanabe, Nicolas Großmann, Christoph Maerz, Shoya Ishimaru, and Andreas Dengel. 2026. Knowledge Transfer with AI. InThe Future of Education with AI: Communications of NII Shonan Meetings. Springer, 51– 86

work page 2026
[27]

Victoria Weiss and Robert Needlman. 1998. To teach is to learn twice: resident teachers learn more.Archives of pediatrics & adolescent medicine 152, 2 (1998), 190–192

work page 1998
[28]

1988.Peer Teaching: To Teach Is To Learn Twice

Neal A Whitman and Jonathan D Fife. 1988.Peer Teaching: To Teach Is To Learn Twice. ASHE-ERIC Higher Education Report No. 4, 1988.ERIC

work page 1988
[29]

Kanta Yamaoka, Ko Watanabe, Koichi Kise, Andreas Dengel, and Shoya Ishimaru. 2023. Experience is the Best Teacher: Personalized Vocabulary Building Within the Context of Instagram Posts and Sentences from GPT-

work page 2023
[30]

Association for Computing Machinery, New York, NY, USA, 313–316

InAdjunct Proceedings of the 2022 ACM International Joint Conference on Pervasive and Ubiquitous Computing and the 2022 ACM International Symposium on Wearable Computers(Cambridge, United Kingdom)(Ubi- Comp/ISWC ’22 Adjunct). Association for Computing Machinery, New York, NY, USA, 313–316. https://doi.org/10.1145/3544793.3560382

work page doi:10.1145/3544793.3560382 2022
[31]

Kanta Yamaoka, Ko Watanabe, Koichi Kise, Andreas Dengel, and Shoya Ishimaru. 2025. Img2Vocab: Explore Words Tied to Your Life with LLMs and Social Media Images.IEEE Access(2025), 1–1. https://doi.org/10.1109/ ACCESS.2025.3533076

work page arXiv 2025
[32]

title": Please follow the format below: Misuse of the

Fangfang Zhu, Jiumin Yang, and Zhongling Pi. 2022. Benefits of Peer Learn- ing and Learning by Teaching for Students Learning through Instructional Videos. In2022 IEEE 2nd International Conference on Educational Technology (ICET). 96–100. https://doi.org/10.1109/ICET55642.2022.9944478 AHs 2026, March 16–19, 2026, Okinawa, Japan Uchida and Watanabe et al. ...

work page doi:10.1109/icet55642.2022.9944478 2022

[1] [2]

Eleanor R Bowyer and Sebastian CK Shaw. 2021. Informal near-peer teaching in medical education: a scoping review.Education for Health34, 1 (2021), 29–33

work page 2021

[2] [3]

HinMingFrankie Chik. 2021. Liji (Book of Rites). https://doi.org/10.14288/ 1.0404466

work page 2021

[3] [4]

2014.The Book of Rites (Liji): Bilingual Edition, English and Chinese

Confucius (Attributed). 2014.The Book of Rites (Liji): Bilingual Edition, English and Chinese. James Legge. https://www.amazon.de/Book-Rites- Liji-Bilingual-English-ebook/dp/B00KVGYS9M Bilingual Edition, English and Chinese

work page 2014

[4] [5]

Claudio G. Cortese. 2005. Learning through Teaching.Management Learn- ing36, 1 (2005), 87–115. https://doi.org/10.1177/1350507605049905

work page doi:10.1177/1350507605049905 2005

[5] [6]

Amy Debbané, Ken Jen Lee, Jarvis Tse, and Edith Law. 2023. Learning by Teaching: Key Challenges and Design Implications.Proc. ACM Hum.- Comput. Interact.7, CSCW1 (April 2023), 1–34. https://doi.org/10.1145/ 3579501

work page 2023

[6] [7]

Jiexin Ding, Bowen Zhao, Yuqi Huang, Yuntao Wang, and Yuanchun Shi

work page

[7] [8]

InExtended Abstracts of the 2023 CHI Conference on Human Factors in Computing Systems(Hamburg, Germany) (CHI EA ’23)

GazeReader: Detecting Unknown Word Using Webcam for English as a Second Language (ESL) Learners. InExtended Abstracts of the 2023 CHI Conference on Human Factors in Computing Systems(Hamburg, Germany) (CHI EA ’23). Association for Computing Machinery, New York, NY, USA, Article 149, 7 pages. https://doi.org/10.1145/3544549.3585790

work page doi:10.1145/3544549.3585790 2023

[8] [9]

Min Fan, Sheng Jin, and Alissa N. Antle. 2018. Designing Colours and Materials in Tangible Reading Products for Foreign Language Learners of English. InExtended Abstracts of the 2018 CHI Conference on Human Factors Empowering Vocabulary Learning Through Teaching AI: Using LLMs as a Student to Perform Learning by Teaching in Vocabulary Acquisition AHs 2026...

work page arXiv 2018

[9] [10]

Logan Fiorella and Richard E. Mayer. 2013. The relative benefits of learning by teaching and teaching expectancy.Contemporary Educational Psychol- ogy38, 4 (2013), 281–288. https://doi.org/10.1016/j.cedpsych.2013.06.001

work page doi:10.1016/j.cedpsych.2013.06.001 2013

[10] [11]

R. C. Gardner and P. D. MacIntyre. 1991. An Instrumental Motivation In Language Study: Who Says It Isn’t Effective?Studies in Second Language Acquisition13, 1 (1991), 57–72. https://doi.org/10.1017/S0272263100009724

work page doi:10.1017/s0272263100009724 1991

[11] [12]

Taichi Higasa, Keitaro Tanaka, Qi Feng, and Shigeo Morishima. 2024. Keep Eyes on the Sentence: An Interactive Sentence Simplification System for English Learners Based on Eye Tracking and Large Language Models. In Extended Abstracts of the CHI Conference on Human Factors in Computing Systems(Honolulu, HI, USA)(CHI EA ’24). Association for Computing Machin...

work page arXiv 2024

[12] [13]

Riku Higashimura, Ko Watanabe, Andrew Vargo, Motoi Iwata, Andreas Dengel, and Koichi Kise. 2024. Estimating Unknown English Words From User Smartphone Reading Behaviors.IEEE Access12 (2024), 140223–140234. https://doi.org/10.1109/ACCESS.2024.3457510

work page doi:10.1109/access.2024.3457510 2024

[13] [14]

Hyoungwook Jin, Seonghee Lee, Hyungyu Shin, and Juho Kim. 2024. Teach AI How to Code: Using Large Language Models as Teachable Agents for Programming Education. InProceedings of the 2024 CHI Conference on Human Factors in Computing Systems (CHI ’24). Association for Computing Machinery, New York, NY, USA, 1—-28. https://doi.org/10.1145/3613904. 3642349

work page doi:10.1145/3613904 2024

[14] [15]

Nayoung Jin and Hana Lee. 2022. StuBot: Learning by Teaching a Con- versational Agent Through Machine Reading Comprehension. InFind- ings of the Association for Computational Linguistics: EMNLP 2022, Yoav Goldberg, Zornitsa Kozareva, and Yue Zhang (Eds.). Association for Com- putational Linguistics, Abu Dhabi, United Arab Emirates, 3008–3020. https://doi....

work page doi:10.18653/v1/2022.findings-emnlp.219 2022

[15] [16]

Pantasdo, Jessy Ceha, Sangho Suh, and Nicole Dillen

Edith Law, Parastoo Baghaei Ravari, Nalin Chhibber, Dana Kulic, Stephanie Lin, Kevin D. Pantasdo, Jessy Ceha, Sangho Suh, and Nicole Dillen. 2020. Curiosity Notebook: A Platform for Learning by Teaching Conversational Agents. InExtended Abstracts of the 2020 CHI Conference on Human Factors in Computing Systems (CHI EA ’20). Association for Computing Machi...

work page doi:10.1145/3334480.3382783 2020

[16] [17]

Ziyi Liu, Zhengzhe Zhu, Lijun Zhu, Enze Jiang, Xiyun Hu, Kylie A Pep- pler, and Karthik Ramani. 2024. ClassMeta: Designing Interactive Vir- tual Classmate to Promote VR Classroom Participation. InProceedings of the 2024 CHI Conference on Human Factors in Computing Systems (CHI ’24). Association for Computing Machinery, New York, NY, USA, 1—-17. https://do...

work page doi:10.1145/3613904.3642947 2024

[17] [18]

Ali Malik, Juliette Woodrow, and Chris Piech. 2024. Learners Teaching Novices: An Uplifting Alternative Assessment. InProceedings of the 55th ACM Technical Symposium on Computer Science Education V. 1 (SIGCSE 2024). Association for Computing Machinery, New York, NY, USA, 785—-

work page 2024

[18] [19]

https://doi.org/10.1145/3626252.3630887

work page doi:10.1145/3626252.3630887

[19] [20]

Markel, Steven G

Julia M. Markel, Steven G. Opferman, James A. Landay, and Chris Piech

work page

[20] [21]

In Proceedings of the Tenth ACM Conference on Learning @ Scale (L@S ’23)

GPTeach: Interactive TA Training with GPT-based Students. In Proceedings of the Tenth ACM Conference on Learning @ Scale (L@S ’23). Association for Computing Machinery, New York, NY, USA, 226—-236. https://doi.org/10.1145/3573051.3593393

work page doi:10.1145/3573051.3593393

[21] [22]

Noboru Matsuda. 2022. Teachable Agent as an Interactive Tool for Cog- nitive Task Analysis: A Case Study for Authoring an Expert Model.In- ternational Journal of Artificial Intelligence in Education32 (2022), 48–75. https://doi.org/10.1007/s40593-021-00265-z

work page doi:10.1007/s40593-021-00265-z 2022

[22] [23]

Maximiliano Paredes-Velasco, Isaac Lozano-Osorio, Diana Pérez-Marín, and Liliana Patricia Santacruz-Valencia. 2024. A Case Study on Learn- ing Visual Programming With TutoApp for Composition of Tutorials: An Approach for Learning by Teaching.IEEE Transactions on Learning Technologies17 (2024), 498–513. https://doi.org/10.1109/TLT.2022.3226122

work page doi:10.1109/tlt.2022.3226122 2024

[23] [24]

Nihar Sabnis and Tomohiro Nagashima. 2024. Empowering Learners: Chatbot-Mediated ’Learning-by-Teaching’. InExtended Abstracts of the CHI Conference on Human Factors in Computing Systems (CHI EA ’24). Association for Computing Machinery, New York, NY, USA, 1—-9. https: //doi.org/10.1145/3613905.3650754

work page doi:10.1145/3613905.3650754 2024

[24] [25]

Fuxing Wang, Meixia Cheng, and Richard Mayer. 2023. Improving learning- by-teaching without audience interaction as a generative learning activity by minimizing the social presence of the audience.Journal of Educational Psychology115, 6 (2023), 783–797. https://doi.org/10.1037/edu0000801

work page doi:10.1037/edu0000801 2023

[25] [26]

Ko Watanabe, Nicolas Großmann, Christoph Maerz, Shoya Ishimaru, and Andreas Dengel. 2026. Knowledge Transfer with AI. InThe Future of Education with AI: Communications of NII Shonan Meetings. Springer, 51– 86

work page 2026

[26] [27]

Victoria Weiss and Robert Needlman. 1998. To teach is to learn twice: resident teachers learn more.Archives of pediatrics & adolescent medicine 152, 2 (1998), 190–192

work page 1998

[27] [28]

1988.Peer Teaching: To Teach Is To Learn Twice

Neal A Whitman and Jonathan D Fife. 1988.Peer Teaching: To Teach Is To Learn Twice. ASHE-ERIC Higher Education Report No. 4, 1988.ERIC

work page 1988

[28] [29]

Kanta Yamaoka, Ko Watanabe, Koichi Kise, Andreas Dengel, and Shoya Ishimaru. 2023. Experience is the Best Teacher: Personalized Vocabulary Building Within the Context of Instagram Posts and Sentences from GPT-

work page 2023

[29] [30]

Association for Computing Machinery, New York, NY, USA, 313–316

InAdjunct Proceedings of the 2022 ACM International Joint Conference on Pervasive and Ubiquitous Computing and the 2022 ACM International Symposium on Wearable Computers(Cambridge, United Kingdom)(Ubi- Comp/ISWC ’22 Adjunct). Association for Computing Machinery, New York, NY, USA, 313–316. https://doi.org/10.1145/3544793.3560382

work page doi:10.1145/3544793.3560382 2022

[30] [31]

Kanta Yamaoka, Ko Watanabe, Koichi Kise, Andreas Dengel, and Shoya Ishimaru. 2025. Img2Vocab: Explore Words Tied to Your Life with LLMs and Social Media Images.IEEE Access(2025), 1–1. https://doi.org/10.1109/ ACCESS.2025.3533076

work page arXiv 2025

[31] [32]

title": Please follow the format below: Misuse of the

Fangfang Zhu, Jiumin Yang, and Zhongling Pi. 2022. Benefits of Peer Learn- ing and Learning by Teaching for Students Learning through Instructional Videos. In2022 IEEE 2nd International Conference on Educational Technology (ICET). 96–100. https://doi.org/10.1109/ICET55642.2022.9944478 AHs 2026, March 16–19, 2026, Okinawa, Japan Uchida and Watanabe et al. ...

work page doi:10.1109/icet55642.2022.9944478 2022