Pith. sign in

REVIEW 3 major objections 5 minor 1 cited by

Classroom Simulacra: Building Contextual Student Generative Agents in Online Education for Learning Behavioral Simulation

T0 review · 3 major / 5 minor · reviewed 2026-08-09 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read A transferable iterative reflection module lets LLM-based student agents predict future quiz correctness more accurately than classical deep learning knowledge tracing models.

desk verdict A useful new dataset and a plausible reflection-based method, but the headline accuracy gain over deep learning baselines is not yet protected against split and retrieval variance. read the letter →

arxiv 2502.02780 v1 pith:E6J62DOR submitted 2025-02-04 cs.HC cs.AIcs.LG

Songlin Xu , Hao-Ning Wen , Hongyi Pan , Dallas Dominguez , Dongyin Hu , Xinyu Zhang This is my paper · ORCID
classification cs.HCcs.AIcs.LG
keywords studentsimulationgenerativeagentsknowledgetracingtransferableiterativereflectionlargelanguagemodelsdigitaltwinclassroomonlineeducationcoursematerials
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

This paper tries to show that large language models, guided by a transferable iterative reflection (TIR) module, can simulate individual students' learning behavior—specifically their future question-answer correctness—more accurately than classical deep learning knowledge tracing models. The motivating problem is that course materials modulate how students learn, yet most student simulation ignores them because datasets lack fine-grained materials and models cannot process long text. The authors collect a six-week, 12-lecture dataset from 60 students with detailed course-material annotations, and they build TIR to distill successful reflection examples from training students so that test-time simulation works even with limited demonstrations. If correct, the result matters because an accurate digital twin classroom would let instructors test pedagogies, personalize learning, and allocate teaching resources without experimenting on real students.

What carries the argument

The central mechanism is the transferable iterative reflection (TIR) module: a two-agent loop in which a reflective LLM explains its wrong predictions on training students, a novice agent re-predicts using that reflection, and the loop continues until 100% accuracy or the maximum iteration count, at which point the best reflection per lecture is logged into a reflection database. At test time, the database supplies example demonstrations from M=4 randomly selected training students in the same lecture, letting a new reflective agent generate one transferable reflection for the test student without seeing ground truth. This compresses long course materials and past history into actionable reasoning, overcoming the token limits of models such as BERT and making prompting-based LLMs more data-efficient.

What would settle it

Hold the model and data fixed but replace the test-time reflections with reflections harvested from a different lecture than the one being simulated; the claim predicts accuracy should fall well below the reported 0.7012 on EduAgent, so little or no drop would indicate that lecture-specific transfer is not the mechanism at work. A second check is to set the number of retrieved training students M to zero and observe whether accuracy falls back toward the non-TIR baseline.

Watch

Extended reading notes

Core claim

The paper's central claim is that TIR lets LLM-based student simulation outperform classical deep learning knowledge tracing models at predicting future question-answer correctness, even with limited demonstration data. The decisive comparison is on the EduAgent dataset: BertKT+TIR reaches 0.7012 accuracy versus SimpleKT's 0.6772, and TIR shows a similar or larger margin on the authors' newly collected dataset. TIR works by running an iterative reflection protocol during training—a reflective agent generates reasons for wrong predictions, a novice agent tests them, and the best reflection per lecture is stored—then, at test time, a new reflective agent retrieves reflections from M=4 randomly selected same-lecture training students and produces a single reflection for the unseen student without ground-truth labels. The paper additionally argues that TIR captures the granular dynamism of learning performance—individual, lecture, question, and skill levels—and inter-student correlations better than non-TIR baselines.

Load-bearing premise

The whole gain rests on the assumption that summaries of why a model made wrong predictions for a few randomly chosen training students in the same lecture can be reused for a new student, even though the new student's summary is written without being told the correct answers; if those summaries are tuned to particular students or lectures rather than generalizable, the accuracy improvement disappears.

Editorial extensions

If this is right

  • TIR-augmented LLMs, including smaller models, can beat classical deep learning knowledge tracing baselines that train on all training students, using only a handful of example demonstrations.
  • The same reflection database can augment prompting-based models (standard and chain-of-thought) and finetuning-based BERT, because reflections compress long course text into short inputs.
  • Simulated students track real students' average accuracy across lectures, questions, and slides; in the authors' data, TIR raised the Pearson correlation on individual differences from 0.02 to 0.42 and on question-level accuracy from -0.50 to 0.37.
  • The framework points toward a digital twin classroom, where instructors can preview how new course materials and test questions would affect a cohort before teaching real students.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • The paper explicitly leaves cross-lecture generalization untested; a direct test would be to build the reflection database from one lecture and apply it to a different lecture, predicting a large accuracy drop if the reflection content is lecture-specific.
  • Because test-time reflections are generated without ground-truth feedback, the method's success likely depends on how well the M=4 randomly selected training reflections transfer; ablating M and switching from random to similarity-based retrieval would reveal how sensitive the gain is.
  • If the mechanism generalizes, a practical consequence is that educators could use cohort-level simulated accuracy to compare curriculum versions before deployment, not just to predict individual quiz scores; a minimal test would compare simulated aggregate accuracy against a real cohort taught with the same new materials.
  • The current pipeline converts slide images to text and excludes lecturer transcripts, so richer material representations could either further improve simulation or show that text-only course content is sufficient for the observed gains.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

3 major / 5 minor

Summary. The paper introduces Classroom Simulacra, a framework for simulating student learning behavior in online education using LLM-based generative agents. The authors contribute a new fine-grained dataset (CogEdu) collected from a 6-week online workshop with 60 students, and a transferable iterative reflection (TIR) module that augments both prompting-based and finetuning-based LLM student simulators. TIR iteratively refines reflections on training students using ground-truth labels, stores successful reflections in a database, and at test time retrieves reflections from M=4 randomly selected training students in the same lecture to guide a reflective agent. The paper evaluates the approach on the EduAgent public dataset and the new CogEdu dataset, comparing against deep learning knowledge tracing baselines (DKT, AKT, ATKT, DKVMN, SimpleKT). The central claim is that TIR enables LLM-based simulation to outperform classical deep learning models in predicting future question-answer correctness, even with limited demonstration data.

Significance. If the central claim holds, this is a valuable contribution to HCI and educational technology. The paper provides a new dataset with granular course-material annotations, which is rare and useful for contextual student modeling. The TIR method is a practical technique for compressing long course materials and transferring reflections across students, and the authors release both the system and model implementations. The paper also goes beyond aggregate accuracy by examining individual-, lecture-, question-, and skill-level correlations, which is a useful evaluation lens. However, the empirical superiority claim rests on a single data split and a single stochastic instantiation of the retrieval procedure, so the significance of the headline result is not yet firmly established; the evaluation needs additional variance analysis and robustness checks before the claim can be fully accepted.

major comments (3)
  1. [Section 3.4, Table 1, Figure 9]
  2. [Section 6.2, Figures 10-12]
  3. [Section 3.4, Section 6.8, Section 7.2.3]
minor comments (5)
  1. [Section 3.4]
  2. [Section 3.2]
  3. [Table 1]
  4. [Section 6.2]
  5. [Figure 9]

Circularity Check

0 steps flagged · score 1.0 of 10

No significant circularity: TIR is a supervised reflection-selection method with a clean train/test split, and the central empirical claim does not reduce to its inputs.

full rationale

The derivation is self-contained with respect to circularity. TIR's training phase (Sec. 3.3) uses ground-truth labels to select the reflection r_best = r_argmax(acc_k) that maximizes a novice agent's accuracy on each training student; this is a supervised prompt-construction procedure, not a prediction. The test phase (Sec. 3.4) explicitly prevents label leakage: 'we use a new reflective agent ... to retrieve reflections from the successful reflection database' and 'this new reflective agent does not experience any other training data,' and the retrieved reflections are used only as demonstrations to generate a fresh reflection for the test student. Test labels enter only as the evaluation target, so the reported accuracies (Table 1, Fig. 9) are not equivalent to the fitted selection objective. The comparison with deep-learning knowledge-tracing baselines uses the same individual-wise train/test split and the same BERT text embeddings (Secs. 3.7, 6.1), so the 'TIR beats DL' claim is an empirical result rather than a definitional consequence. The only self-citations are to the authors' own public EduAgent dataset [86] and prior generative-agent work [85]; EduAgent is used as a public benchmark alongside the newly collected CogEdu dataset, and no theoretical premise or ansatz is imported from these citations. Concerns about a single split, small test sets (df=17), and unreported retrieval seed/variance are correctness/reproducibility risks, not circularity.

Assumptions & free parameters 4 free parameters · 5 assumptions · 0 invented entities

The central results rest on assumptions about LLM reflection quality, transferability across students, and the validity of quiz accuracy as a behavioral proxy. The main free parameters are the number of retrieved reflections M and the number of past questions; both are chosen without sensitivity analysis.

free parameters (4)
  • M (number of retrieved reflections) = 4
    Hand-picked based on a pilot experiment (Section 3.4); affects test-time prompt content and results.
  • Number of past questions = 5
    Arbitrary choice in Section 3.2; defines the learning history input.
  • Train/test split ratio = 0.7 for CogEdu, 0.8 for EduAgent
    Individual-wise split; no multiple splits or error bars reported.
  • Maximum TIR iterations = not specified
    Stopping condition in Section 3.3; iteration count affects reflection quality but is not reported.
assumptions (5)
  • domain assumption LLMs can generate useful self-reflections about prediction errors when given ground-truth labels.
    Section 3.3 relies on this to construct the reflection database.
  • domain assumption Reflections distilled from training students generalize to test students in the same lecture.
    Section 3.4 assumes transferability; the paper notes it has not tested cross-lecture generalization.
  • domain assumption Course materials can be represented as text; image content is converted to textual descriptions.
    Section 3.1 states slides are converted to text; this ignores visual and spatial information.
  • domain assumption Post-test accuracy is a sufficient proxy for student learning behavior.
    The evaluation uses only binary correctness; limitations section acknowledges other behaviors are not modeled.
  • domain assumption The first five questions provide enough history to simulate future performance.
    Section 3.2 sets the first five questions as past; this truncates the learning history.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Classroom Simulacra: Building Contextual Student Generative Agents in Online Education for Learning Behavioral Simulation." pith.science (2026). https://pith.science/paper/E6J62DOR

@misc{pith2026250202780,
  author       = {Pith},
  title        = {Pith review of: Classroom Simulacra: Building Contextual Student Generative Agents in Online Education for Learning Behavioral Simulation},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/E6J62DOR}},
  note         = {Machine review of arXiv:2502.02780}
}
read the original abstract

Student simulation supports educators to improve teaching by interacting with virtual students. However, most existing approaches ignore the modulation effects of course materials because of two challenges: the lack of datasets with granularly annotated course materials, and the limitation of existing simulation models in processing extremely long textual data. To solve the challenges, we first run a 6-week education workshop from N = 60 students to collect fine-grained data using a custom built online education system, which logs students' learning behaviors as they interact with lecture materials over time. Second, we propose a transferable iterative reflection (TIR) module that augments both prompting-based and finetuning-based large language models (LLMs) for simulating learning behaviors. Our comprehensive experiments show that TIR enables the LLMs to perform more accurate student simulation than classical deep learning models, even with limited demonstration data. Our TIR approach better captures the granular dynamism of learning performance and inter-student correlations in classrooms, paving the way towards a ''digital twin'' for online education.

Figures

Figures reproduced from arXiv: 2502.02780 by the authors.

Figure 1
Figure 1. Overview of our Classroom Simulacra framework which uses a transferable iterative reflection method to effectively [PITH_FULL_IMAGE:figures/full_fig_p001_1.png] view at source ↗
Figure 2
Figure 2. Training (left) and testing (right) schemes for prompting-based models. 3 Simulation Methodology Our classroom simulacra framework aims to build LLM-based gen￾erative student agents that could mimic real students’ learning behaviors based on their learning histories. The agents can then simulate the students’ future learning performance, which is repre￾sented by the question answering correctness in the post-lecture… view at source ↗
Figure 3
Figure 3. Training (left) and testing (right) scheme for finetuning-based models. its previous simulations by comparing with labels in the example demonstrations (i.e. past learning history) so that LLMs could gen￾erate general reflections results that could be easily transferred to novice LLMs which do not have the example demonstrations. This ensures the generalization of such reflected results to be applied into new future… view at source ↗
Figures from the paper (15 more)
Figure 4
Figure 4. Figure 4: Prompt examples in the Transferable Iterative Reflection process. and 𝑙 𝑓 𝑢𝑡𝑢𝑟𝑒 to deal with the token limit problem. More details are described in the following sections. 3.4 TIR Augments Standard Prompts Here we describe how to apply the TIR module to augment the sta…
Figure 5
Figure 5. Figure 5: (a). Illustration of our CogEdu system. (b). Our ac￾tion prompt strategy for instructors based on attention ratio and knowledge ratio. (c,d). The UI of the user end (c) and server end (d) of CogEdu. As depicted in Section 3, our TIR module could augment both prompting-…
Figure 7
Figure 7. Figure 7: The procedure to use our online learning system. (a)(b)(c). Gaze calibration process for gaze tracking. (d)(e). Fa￾cial expression model training data collection process. (f)(g). Students and teachers join in the online video calling from their own clients in class. te…
Figure 8
Figure 8. Figure 8: (a)(b). BertKT training curve without (a) and with (b) TIR across different epochs. Vertical axis on the left is the loss value (solid lines). Vertical axis on the right is the metrics (accuracy and F1 score) value (dotted lines). communication between students and ins…
Figure 9
Figure 9. Figure 9: Model accuracy and F1 score comparison on our newly collected dataset. Left (a,c) shows comparison (a: accuracy, c: F1 score) among deep learning models and LLMs-based models using GPT-4o. Right (b,d) shows comparison (b: accuracy, d: F1 score) of LLMs-based models usi…
Figure 10
Figure 10. Figure 10: Heatmap to show the average simulation accuracy (each cell) for each individual student using each model. adaptive learning strategies [66]. Therefore, the superior predic￾tive capability of our model indicates a strong potential for real￾world applicability. To furth…
Figure 11
Figure 11. Figure 11: Heatmap to show the average simulation accuracy (each cell) for each specific lecture using each model. increased all of the GPT-4o mini based simulation models (Standard, CoT, BertKT), which were even better than the simulation models using GPT-4o without the TIR mod…
Figure 13
Figure 13. Figure 13: Individual-level results: simulations using BertKT with and without TIR across different students. (a). Correlation between student ID and simulated/real post-test question answer accuracy (vertical bar: standard deviation). (b,c,d,e). The distribution of simulation a…
Figure 14
Figure 14. Figure 14: Lecture-level results: simulations using BertKT with and without TIR across different lecture IDs. (a). Correlation between lecture ID and simulated/real post-test question answer accuracy (vertical bar: standard deviation). (b)(c)(d). Heatmaps of label, BertKT with T…
Figure 15
Figure 15. Figure 15: Question-level results: simulations using BertKT with and without TIR across different post-test question IDs. (a). Correlation between question ID and simulated/real post-test question answer accuracy (vertical bar: standard deviation). (b)(c)(d). Heatmaps of label, …
Figure 16
Figure 16. Figure 16: (a). Question answer accuracy of simulations using BertKT with and without TIR against the labels on our newly collected dataset across questions related to each slide. Each star point depicts the average question answer accuracy of questions related to that slide whi…
Figure 17
Figure 17. Figure 17: Inter-student correlation graphs of the label (a) and simulations using BertKT with (b) and without TIR (c) on our [PITH_FULL_IMAGE:figures/full_fig_p018_17.png]
Figure 18
Figure 18. Figure 18: Example prompts of standard prompting without (left) and with (right) TIR. Contents are truncated. [PITH_FULL_IMAGE:figures/full_fig_p023_18.png]
Figure 19
Figure 19. Figure 19: Example prompts of CoT prompting without (left) and with (right) TIR. Contents are truncated. [PITH_FULL_IMAGE:figures/full_fig_p024_19.png]
Figure 20
Figure 20. Figure 20: Example prompts of BertKT without (left) and with (right) our TIR module. The input prompts for BertKT are [PITH_FULL_IMAGE:figures/full_fig_p025_20.png]

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 1 Pith paper

Reviewed papers in the Pith corpus that reference this work. Sorted by Pith novelty score. Full citation record

  1. Conversational Education at Scale: A Multi-LLM Agent Workflow for Procedural Learning and Pedagogic Quality Assessment

    cs.AI 2025-07 conditional novelty 5.0 of 10

    WikiHowAgent generates 114,296 simulated teacher-learner conversations from 14,287 WikiHow tutorials and evaluates their pedagogic quality with LLM and human judges.

Reference graph

Works this paper leans on

114 extracted references · 38 canonical work pages · cited by 1 Pith paper

  1. [1]

    Ghodai Abdelrahman, Qing Wang, and Bernardo Nunes. 2023. Knowledge tracing: A survey. Comput. Surveys 55, 11 (2023), 1–37

    2023

  2. [2]

    Badr AlKhamissi, Millicent Li, Asli Celikyilmaz, Mona Diab, and Marjan Ghazvininejad. 2022. A review on language models as knowledge bases. arXiv preprint arXiv:2204.06031 (2022)

    arXiv 2022

  3. [3]

    Riku Arakawa, Hiromu Yakura, and Masataka Goto. 2023. CatAlyst: domain- extensible intervention for preventing task procrastination using large generative models. In Proceedings of the 2023 CHI Conference on Human Factors in Computing Systems. 1–19

    2023

  4. [4]

    Ananya Bhattacharjee, Yuchen Zeng, Sarah Yi Xu, Dana Kulzhabayeva, Minyi Ma, Rachel Kornfield, Syed Ishtiaque Ahmed, Alex Mariakakis, Mary P Czerwinski, Anastasia Kuzminykh, et al. 2024. Understanding the Role of Large Language Models in Personalizing and Scaffolding Strategies to Combat Academic Procras- tination. In Proceedings of the CHI Conference on ...

    2024

  5. [5]

    Paul Calle, Ruosi Shao, Yunlong Liu, Emily T Hébert, Darla Kendzor, Jordan Neil, Michael Businelle, and Chongle Pan. 2024. Towards AI-Driven Healthcare: Systematic Optimization, Linguistic Analysis, and Clinicians’ Evaluation of Large Language Models for Smoking Cessation Interventions. In Proceedings of the CHI Conference on Human Factors in Computing Sy...

    2024

  6. [6]

    Weize Chen, Yusheng Su, Jingwei Zuo, Cheng Yang, Chenfei Yuan, Chen Qian, Chi-Min Chan, Yujia Qin, Yaxi Lu, Ruobing Xie, et al. 2023. Agentverse: Facilitat- ing multi-agent collaboration and exploring emergent behaviors in agents. arXiv preprint arXiv:2308.10848 2, 4 (2023), 6

    arXiv 2023

  7. [7]

    Youngduck Choi, Youngnam Lee, Dongmin Shin, Junghyun Cho, Seoyon Park, Seewoo Lee, Jineon Baek, Chan Bae, Byungsoo Kim, and Jaewe Heo. 2020. Ednet: A large-scale hierarchical dataset in education. In Artificial Intelligence in Edu- cation: 21st International Conference, AIED 2020, Ifrane, Morocco, July 6–10, 2020, Proceedings, Part II 21 . Springer, 69–73

    2020

  8. [8]

    Peng Cui and Mrinmaya Sachan. 2023. Adaptive and personalized exercise generation for online language learning. arXiv preprint arXiv:2306.02457 (2023)

    work page Pith review arXiv 2023

Show all 114 references
  1. [9]

    Yang Deng, An Zhang, Yankai Lin, Xu Chen, Ji-Rong Wen, and Tat-Seng Chua

  2. [10]

    Jacob Devlin, Ming-Wei Chang, Kenton Lee, and Kristina Toutanova. 2018. BERT: Pre-training of Deep Bidirectional Transformers for Language Understanding. CoRR abs/1810.04805 (2018). arXiv:1810.04805 http://arxiv.org/abs/1810.04805

    2018 arXiv

  3. [11]

    Qingxiu Dong, Lei Li, Damai Dai, Ce Zheng, Zhiyong Wu, Baobao Chang, Xu Sun, Jingjing Xu, and Zhifang Sui. 2022. A survey on in-context learning. arXiv preprint arXiv:2301.00234 (2022)

    2022 arXiv

  4. [12]

    Peitong Duan, Jeremy Warner, Yang Li, and Bjoern Hartmann. 2024. Generating Automatic Feedback on UI Mockups with Large Language Models. In Proceedings of the CHI Conference on Human Factors in Computing Systems . 1–20

    2024

  5. [13]

    Ralf Engbert and Reinhold Kliegl. 2003. Microsaccades uncover the orientation of covert attention. Vision research 43, 9 (2003), 1035–1045

    2003

  6. [14]

    Raymond Fok, Nedim Lipka, Tong Sun, and Alexa F Siu. 2024. Marco: Supporting Business Document Workflows via Collection-Centric Information Foraging with Large Language Models. In Proceedings of the CHI Conference on Human Factors in Computing Systems. 1–20

    2024

  7. [15]

    Lingyue Fu, Hao Guan, Kounianhua Du, Jianghao Lin, Wei Xia, Weinan Zhang, Ruiming Tang, Yasheng Wang, and Yong Yu. 2024. SINKT: A Structure- Aware Inductive Knowledge Tracing Model with Large Language Model. arXiv:2407.01245 [cs.AI] https://arxiv.org/abs/2407.01245

    2024 arXiv

  8. [16]

    Jie Gao, Yuchen Guo, Gionnieve Lim, Tianqin Zhang, Zheng Zhang, Toby Jia- Jun Li, and Simon Tangi Perrault. 2024. CollabCoder: a lower-barrier, rigorous workflow for inductive collaborative qualitative analysis with large language models. In Proceedings of the CHI Conference o...

    2024

  9. [17]

    Simret Araya Gebreegziabher, Zheng Zhang, Xiaohang Tang, Yihao Meng, Elena L Glassman, and Toby Jia-Jun Li. 2023. Patat: Human-ai collaborative qualitative coding with explainable interactive rule synthesis. In Proceedings of the 2023 CHI Conference on Human Factors in Computi...

    2023

  10. [18]

    Aritra Ghosh, Neil Heffernan, and Andrew S. Lan. 2020. Context-Aware Attentive Knowledge Tracing. In Proceedings of the 26th ACM SIGKDD International Con- ference on Knowledge Discovery & Data Mining (Virtual Event, CA, USA) (KDD ’20). Association for Computing Machinery, New ...

    2020

  11. [19]

    Arthur C Graesser, Mark W Conley, and Andrew Olney. 2012. Intelligent tutoring systems. (2012)

    2012

  12. [20]

    Yuxian Gu, Li Dong, Furu Wei, and Minlie Huang. 2024. MiniLLM: Knowledge Distillation of Large Language Models. In The Twelfth International Conference on Learning Representations. https://openreview.net/forum?id=5h0qf7IBZZ

    2024

  13. [21]

    Xiaopeng Guo, Zhijie Huang, Jie Gao, Mingyu Shang, Maojing Shu, and Jun Sun

  14. [22]

    Muhammad Usman Hadi, Rizwan Qureshi, Abbas Shah, Muhammad Irfan, Anas Zafar, Muhammad Bilal Shaikh, Naveed Akhtar, Jia Wu, Seyedali Mirjalili, et al

  15. [23]

    Xinying Hou, Zihan Wu, Xu Wang, and Barbara J Ericson. 2024. Codetailor: Llm-powered personalized parsons puzzles for engaging support while learning programming. In Proceedings of the Eleventh ACM Conference on Learning@ Scale . 51–62

    2024

  16. [24]

    Yihan Hou, Manling Yang, Hao Cui, Lei Wang, Jie Xu, and Wei Zeng. 2024. C2Ideas: Supporting Creative Interior Color Design Ideation with a Large Lan- guage Model. InProceedings of the CHI Conference on Human Factors in Computing Systems. 1–18

    2024

  17. [25]

    Wenyang Hui, Yan Wang, Kewei Tu, and Chengyue Jiang. 2024. RoT: Enhanc- ing Large Language Models with Reflection on Search Trees. arXiv preprint arXiv:2404.05449 (2024)

    2024 arXiv

  18. [26]

    It’s the only thing I can trust

    JiWoong Jang, Sanika Moharana, Patrick Carrington, and Andrew Begel. 2024. “It’s the only thing I can trust”: Envisioning Large Language Model Use by Autistic Workers for Communication Assistance. In Proceedings of the CHI Conference on Human Factors in Computing Systems . 1–18

    2024

  19. [27]

    Ziwei Ji, Tiezheng Yu, Yan Xu, Nayeon Lee, Etsuko Ishii, and Pascale Fung. 2023. Towards mitigating hallucination in large language models via self-reflection. arXiv preprint arXiv:2310.06271 (2023)

    2023 arXiv

  20. [28]

    Hyoungwook Jin, Seonghee Lee, Hyungyu Shin, and Juho Kim. 2024. Teach AI How to Code: Using Large Language Models as Teachable Agents for Pro- gramming Education. In Proceedings of the CHI Conference on Human Factors in Computing Systems. 1–28

    2024

  21. [29]

    Eunkyung Jo, Daniel A Epstein, Hyunhoon Jung, and Young-Ho Kim. 2023. Under- standing the benefits and challenges of deploying conversational AI leveraging large language models for public health intervention. In Proceedings of the 2023 CHI Conference on Human Factors in Compu...

    2023

  22. [30]

    Eunkyung Jo, Yuin Jeong, SoHyun Park, Daniel A Epstein, and Young-Ho Kim

  23. [31]

    Heeseok Jung, Jaesang Yoo, Yohaan Yoon, and Yeonju Jang. 2024. CLST: Cold- Start Mitigation in Knowledge Tracing by Aligning a Generative Language Model as a Students’ Knowledge Tracer. arXiv preprint arXiv:2406.10296 (2024)

    2024 arXiv

  24. [32]

    Minsol Kim, Aliea L Nallbani, and Abby Rayne Stovall. 2024. Exploring LLM- based Chatbot for Language Learning and Cultivation of Growth Mindset. In Extended Abstracts of the CHI Conference on Human Factors in Computing Systems . 1–5

    2024

  25. [33]

    Taewan Kim, Seolyeong Bae, Hyun Ah Kim, Su-woo Lee, Hwajung Hong, Chanmo Yang, and Young-Ho Kim. 2024. MindfulDiary: Harnessing Large Language Model to Support Psychiatric Patients’ Journaling. In Proceedings of the CHI Conference on Human Factors in Computing Systems . 1–20

    2024

  26. [34]

    In Pro- ceedings of the CHI Conference on Human Factors in Computing Systems

    Understanding the Impact of Long-Term Memory on Self-Disclosure with Large Language Model-Driven Chatbots for Public Health Intervention. In Pro- ceedings of the CHI Conference on Human Factors in Computing Systems . 1–21

  27. [35]

    Yann LeCun, Yoshua Bengio, and Geoffrey Hinton. 2015. Deep learning. nature 521, 7553 (2015), 436–444

    2015

  28. [36]

    Unggi Lee, Jiyeong Bae, Dohee Kim, Sookbun Lee, Jaekwon Park, Taekyung Ahn, Gunho Lee, Damji Stratton, and Hyeoncheol Kim. 2024. Language Model Can Do Knowledge Tracing: Simple but Effective Method to Integrate Language Model and Knowledge Tracing Task. arXiv preprint arXiv:24...

    2024 arXiv

  29. [37]

    Unggi Lee, Yonghyun Park, Yujin Kim, Seongyune Choi, and Hyeoncheol Kim

  30. [38]

    Harsh Kumar, Ruiwei Xiao, Benjamin Lawson, Ilya Musabirov, Jiakai Shi, Xinyuan Wang, Huayin Luo, Joseph Jay Williams, Anna N Rafferty, John Stamper, et al

  31. [39]

    InProceedings of the Eleventh ACM Conference on Learning@ Scale

    Supporting Self-Reflection at Scale with Large Language Models: Insights from Randomized Field Experiments in Classrooms. InProceedings of the Eleventh ACM Conference on Learning@ Scale . 86–97

  32. [40]

    Qingyao Li, Wei Xia, Kounianhua Du, Qiji Zhang, Weinan Zhang, Ruiming Tang, and Yong Yu. 2024. Learning Structure and Knowledge Aware Representation with Large Language Models for Concept Recommendation. arXiv preprint arXiv:2405.12442 (2024)

    2024 arXiv

  33. [41]

    Tianle Li, Ge Zhang, Quy Duc Do, Xiang Yue, and Wenhu Chen. 2024. Long-context llms struggle with long in-context learning. arXiv preprint arXiv:2404.02060 (2024)

    2024 arXiv

  34. [42]

    Zhaoxing Li, Jujie Yang, Jindi Wang, Lei Shi, and Sebastian Stein. 2024. Integrating lstm and bert for long-sequence data analysis in intelligent tutoring systems. arXiv preprint arXiv:2405.05136 (2024)

    2024 arXiv

  35. [43]

    In International Conference on Intelligent Tutoring Systems

    Monacobert: Monotonic attention based convbert for knowledge tracing. In International Conference on Intelligent Tutoring Systems . Springer, 107–123

  36. [44]

    Haoxuan Li, Jifan Yu, Yuanxin Ouyang, Zhuang Liu, Wenge Rong, Juanzi Li, and Zhang Xiong. 2024. Explainable few-shot knowledge tracing. arXiv preprint arXiv:2405.14391 (2024)

    2024 arXiv

  37. [45]

    Moxin Li, Wenjie Wang, Fuli Feng, Fengbin Zhu, Qifan Wang, and Tat-Seng Chua. 2024. Think twice before assure: Confidence estimation for large language models through reflection on multiple answers. arXiv preprint arXiv:2403.09972 (2024)

    2024 arXiv

  38. [46]

    Zitao Liu, Qiongqiong Liu, Jiahao Chen, Shuyan Huang, and Weiqi Luo. 2023. simpleKT: A Simple But Tough-to-Beat Baseline for Knowledge Tracing. In The Eleventh International Conference on Learning Representations, ICLR 2023, Kigali, Rwanda, May 1-5, 2023 . OpenReview.net. http...

    2023

  39. [47]

    Zitao Liu, Qiongqiong Liu, Jiahao Chen, Shuyan Huang, Jiliang Tang, and Weiqi Luo. 2022. pyKT: A Python Library to Benchmark Deep Learning based Knowledge Tracing Models. In Advances in Neural Infor- mation Processing Systems , S. Koyejo, S. Mohamed, A. Agarwal, D. Bel- grave,...

    2022

  40. [48]

    Helen E Longino. 2019. Studying human behavior: How scientists investigate aggression and sexuality. University of Chicago Press

    2019

  41. [49]

    Zhenwen Liang, Wenhao Yu, Tanmay Rajpurohit, Peter Clark, Xiangliang Zhang, and Ashwin Kaylan. 2023. Let gpt be a math tutor: Teaching math word problem solvers with customized exercise generation. arXiv preprint arXiv:2305.14386 (2023)

    2023 arXiv

  42. [50]

    What it wants me to say

    Michael Xieyang Liu, Advait Sarkar, Carina Negreanu, Benjamin Zorn, Jack Williams, Neil Toronto, and Andrew D Gordon. 2023. “What it wants me to say”: Bridging the abstraction gap between end-user programmers and code- generating large language models. In Proceedings of the 20...

    2023

  43. [51]

    Naiming Liu, Zichao Wang, Richard G Baraniuk, and Andrew Lan. 2022. GPT- based Open-ended Knowledge Tracing. arXiv preprint arXiv:2203.03716 (2022)

    2022 arXiv

  44. [52]

    Aman Madaan, Niket Tandon, Prakhar Gupta, Skyler Hallinan, Luyu Gao, Sarah Wiegreffe, Uri Alon, Nouha Dziri, Shrimai Prabhumoye, Yiming Yang, et al

  45. [53]

    Julia M Markel, Steven G Opferman, James A Landay, and Chris Piech. 2023. Gpteach: Interactive ta training with gpt-based students. In Proceedings of the tenth acm conference on learning@ scale . 226–236

    2023

  46. [54]

    Jordan K Matelsky, Felipe Parodi, Tony Liu, Richard D Lange, and Konrad P Ko- rding. 2023. A large language model-assisted education tool to provide feedback on open-ended responses. arXiv preprint arXiv:2308.02439 (2023)

    2023 arXiv

  47. [55]

    Xinyi Lu, Simin Fan, Jessica Houghton, Lu Wang, and Xu Wang. 2023. Read- ingQuizMaker: a human-NLP collaborative system that supports instructors to design high-quality reading quiz questions. In Proceedings of the 2023 CHI Conference on Human Factors in Computing Systems . 1–18

    2023

  48. [56]

    Xinyi Lu and Xu Wang. 2024. Generative Students: Using LLM-Simulated Student Profiles to Support Question Item Evaluation. In Proceedings of the Eleventh ACM Conference on Learning @ Scale (L@S ’24, Vol. 8). ACM, 16–27. doi:10.1145/3657604. 3662031

    2024 doi

  49. [57]

    Zilin Ma, Yiyang Mei, Yinru Long, Zhaoyuan Su, and Krzysztof Z Gajos. 2024. Evaluating the Experience of LGBTQ+ People Using Large Language Model Based Chatbots for Mental Health Support. In Proceedings of the CHI Conference Classroom Simulacra: Building Contextual Student Gen...

    2024

  50. [58]

    Joon Sung Park, Joseph O’Brien, Carrie Jun Cai, Meredith Ringel Morris, Percy Liang, and Michael S Bernstein. 2023. Generative agents: Interactive simulacra of human behavior. In Proceedings of the 36th annual acm symposium on user interface software and technology . 1–22

    2023

  51. [59]

    Advances in Neural Information Processing Systems 36 (2024)

    Self-refine: Iterative refinement with self-feedback. Advances in Neural Information Processing Systems 36 (2024)

    2024

  52. [60]

    Chris Piech, Jonathan Spencer, Jonathan Huang, Surya Ganguli, Mehran Sahami, Leonidas Guibas, and Jascha Sohl-Dickstein. 2015. Deep Knowledge Tracing. arXiv:1506.05908 [cs.AI] https://arxiv.org/abs/1506.05908

    2015 arXiv

  53. [61]

    Chen Pojen, Hsieh Mingen, and Tsai Tzuyang. 2020. Junyi Academy On- line Learning Activity Dataset: A large-scale public online learning activity dataset from elementary to senior high school students. Dataset available from https://www.kaggle.com/junyiacademy/learning-activit...

    2020

  54. [62]

    Hiromi Nakagawa, Yusuke Iwasawa, and Yutaka Matsuo. 2019. Graph-based knowledge tracing: modeling student proficiency using graph neural network. In IEEE/WIC/ACM International Conference on Web Intelligence . 156–163

    2019

  55. [63]

    Seyed Parsa Neshaei, Richard Lee Davis, Adam Hazimeh, Bojan Lazarevski, Pierre Dillenbourg, and Tanja Käser. 2024. Towards Modeling Learner Performance with Large Language Models. arXiv:2403.14661 [cs.CY] https://arxiv.org/abs/ 2403.14661

    2024 arXiv

  56. [64]

    Tuan Nguyen. 2015. The effectiveness of online learning: Beyond no significant difference and future horizons. MERLOT Journal of online learning and teaching 11, 2 (2015), 309–319

    2015

  57. [65]

    Fatemeh Sarshartehrani, Elham Mohammadrezaei, Majid Behravan, and Denis Gracanin. 2024. Enhancing E-Learning Experience Through Embodied AI Tutors in Immersive Virtual Environments: A Multifaceted Approach for Personalized Educational Adaptation. In International Conference on...

    2024

  58. [66]

    Savvas Petridis, Nicholas Diakopoulos, Kevin Crowston, Mark Hansen, Keren Henderson, Stan Jastrzebski, Jeffrey V Nickerson, and Lydia B Chilton. 2023. Anglekindling: Supporting journalistic angle ideation with large language models. In Proceedings of the 2023 CHI conference on...

    2023

  59. [67]

    Woosuk Seo, Chanmo Yang, and Young-Ho Kim. 2024. ChaCha: Leveraging Large Language Models to Prompt Children to Share Their Emotions about Personal Events. In Proceedings of the CHI Conference on Human Factors in Computing Systems. 1–20

    2024

  60. [68]

    Konstantin R Strömel, Stanislas Henry, Tim Johansson, Jasmin Niess, and Paweł W Woźniak. 2024. Narrating Fitness: Leveraging Large Language Models for Reflec- tive Fitness Tracker Data Interpretation. In Proceedings of the CHI Conference on Human Factors in Computing Systems . 1–16

    2024

  61. [69]

    Hua Xuan Qin, Shan Jin, Ze Gao, Mingming Fan, and Pan Hui. 2024. Char- acterMeet: Supporting Creative Writers’ Entire Story Character Construction Processes Through Conversation with LLM-Powered Chatbot Avatars. In Pro- ceedings of the CHI Conference on Human Factors in Comput...

    2024

  62. [70]

    Niroop Channa Rajashekar, Yeo Eun Shin, Yuan Pu, Sunny Chung, Kisung You, Mauro Giuffre, Colleen E Chan, Theo Saarinen, Allen Hsiao, Jasjeet Sekhon, et al

  63. [71]

    In Proceedings of the CHI Conference on Human Factors in Computing Systems

    Human-Algorithmic Interaction Using a Large Language Model-Augmented Artificial Intelligence Clinical Decision Support System. In Proceedings of the CHI Conference on Human Factors in Computing Systems . 1–20

  64. [72]

    Mohi Reza, Nathan M Laundry, Ilya Musabirov, Peter Dushniku, Zhi Yuan “Michael” Yu, Kashish Mittal, Tovi Grossman, Michael Liut, Anastasia Kuzminykh, and Joseph Jay Williams. 2024. ABScribe: Rapid Exploration & Or- ganization of Multiple Writing Variations in Human-AI Co-Writi...

    2024

  65. [73]

    Bryan Wang, Gang Li, and Yang Li. 2023. Enabling conversational interaction with mobile ui using large language models. In Proceedings of the 2023 CHI Conference on Human Factors in Computing Systems . 1–17

    2023

  66. [74]

    Frances Scholtz and Suzaan Hughes. 2021. A systematic review of educator interventions in facilitating simulation based learning.Journal of Applied Research in Higher Education 13, 5 (2021), 1408–1435

    2021

  67. [75]

    Tianjia Wang, Ramaraja Ramanujan, Yi Lu, Chenyu Mao, Yan Chen, and Chris Brown. 2024. DevCoach: Supporting Students in Learning the Software Develop- ment Life Cycle at Scale with Generative Agents. In Proceedings of the Eleventh ACM Conference on Learning@ Scale . 351–355

    2024

  68. [76]

    Wei Wang, Lihuan Guo, Ling He, and Yenchun Jim Wu. 2019. Effects of social- interactive engagement on the dropout ratio in online learning: insights from MOOC. Behaviour & Information Technology 38, 6 (2019), 621–636

    2019

  69. [77]

    Sangho Suh, Meng Chen, Bryan Min, Toby Jia-Jun Li, and Haijun Xia. 2024. Luminate: Structured Generation and Exploration of Design Space with Large Language Models for Human-AI Co-Creation. InProceedings of the CHI Conference on Human Factors in Computing Systems . 1–26

    2024

  70. [78]

    Xiangru Tang, Yiming Zong, Yilun Zhao, Arman Cohan, and Mark Gerstein. 2023. Struc-Bench: Are Large Language Models Really Good at Generating Complex Structured Data? arXiv preprint arXiv:2309.08963 (2023)

    2023 arXiv

  71. [79]

    Annapurna Vadaparty, Daniel Zingaro, David H Smith IV, Mounika Padala, Chris- tine Alvarado, Jamie Gorson Benario, and Leo Porter. 2024. CS1-LLM: Integrating LLMs into CS1 Instruction. InProceedings of the 2024 on Innovation and Technology in Computer Science Education V. 1 . 297–303

    2024

  72. [80]

    Hongyu Wan, Jinda Zhang, Abdulaziz Arif Suria, Bingsheng Yao, Dakuo Wang, Yvonne Coady, and Mirjana Prpa. 2024. Building LLM-based AI Agents in Social Virtual Reality. In Extended Abstracts of the CHI Conference on Human Factors in Computing Systems. 1–7

    2024

  73. [81]

    Jerry Wei, Jason Wei, Yi Tay, Dustin Tran, Albert Webson, Yifeng Lu, Xinyun Chen, Hanxiao Liu, Da Huang, Denny Zhou, et al. 2023. Larger language models do in-context learning differently. arXiv preprint arXiv:2303.03846 (2023)

    2023 arXiv

  74. [82]

    Sitong Wang, Savvas Petridis, Taeahn Kwon, Xiaojuan Ma, and Lydia B Chilton

  75. [83]

    In Proceedings of the 2023 CHI Conference on Human Factors in Computing Systems

    PopBlends: Strategies for conceptual blending with large language models. In Proceedings of the 2023 CHI Conference on Human Factors in Computing Systems . 1–19

    2023

  76. [84]

    Wanli Xing and Dongping Du. 2019. Dropout prediction in MOOCs: Using deep learning for personalized intervention. Journal of Educational Computing Research 57, 3 (2019), 547–570

    2019

  77. [85]

    Songlin Xu and Xinyu Zhang. 2023. Leveraging generative artificial intelligence to simulate student learning behavior. arXiv preprint arXiv:2310.19206 (2023)

    2023 arXiv

  78. [86]

    Xinru Wang, Hannah Kim, Sajjadur Rahman, Kushan Mitra, and Zhengjie Miao

  79. [87]

    In Proceedings of the CHI Conference on Human Factors in Computing Systems

    Human-LLM collaborative annotation through effective verification of LLM labels. In Proceedings of the CHI Conference on Human Factors in Computing Systems. 1–21

  80. [88]

    Yutong Wang, Jiali Zeng, Xuebo Liu, Fandong Meng, Jie Zhou, and Min Zhang

  81. [89]

    arXiv preprint arXiv:2406.08434 (2024)

    TasTe: Teaching Large Language Models to Translate through Self- Reflection. arXiv preprint arXiv:2406.08434 (2024)

    2024 arXiv

  82. [90]

    Zhan Wang, Lin-Ping Yuan, Liangwei Wang, Bingchuan Jiang, and Wei Zeng

  83. [91]

    In Proceedings of the CHI conference on human factors in computing systems

    Virtuwander: Enhancing multi-modal interaction for virtual tour guidance through large language models. In Proceedings of the CHI conference on human factors in computing systems . 1–20

  84. [92]

    Jason Wei, Xuezhi Wang, Dale Schuurmans, Maarten Bosma, Brian Ichter, Fei Xia, Ed Chi, Quoc Le, and Denny Zhou. 2023. Chain-of-Thought Prompting Elicits Reasoning in Large Language Models. arXiv:2201.11903 [cs.CL] https: //arxiv.org/abs/2201.11903

    2023 arXiv

  85. [93]

    Chao Zhang, Xuechen Liu, Katherine Ziska, Soobin Jeon, Chi-Lin Yu, and Ying Xu. 2024. Mathemyths: leveraging large language models to teach mathematical language through Child-AI co-creative storytelling. In Proceedings of the CHI Conference on Human Factors in Computing Syste...

    2024

  86. [94]

    Ruolan Wu, Chun Yu, Xiaole Pan, Yujia Liu, Ningning Zhang, Yue Fu, Yuhan Wang, Zhi Zheng, Li Chen, Qiaolei Jiang, et al . 2024. MindShift: Leveraging Large Language Models for Mental-States-Based Problematic Smartphone Use Intervention. In Proceedings of the CHI Conference on ...

    2024

  87. [95]

    Tongshuang Wu, Michael Terry, and Carrie Jun Cai. 2022. Ai chains: Transparent and controllable human-ai interaction by chaining large language model prompts. In Proceedings of the 2022 CHI conference on human factors in computing systems . 1–22

    2022

  88. [96]

    Zheyuan Zhang, Daniel Zhang-Li, Jifan Yu, Linlu Gong, Jinchang Zhou, Zhiyuan Liu, Lei Hou, and Juanzi Li. 2024. Simulating classroom education with llm- empowered agents. arXiv preprint arXiv:2406.19226 (2024)

    2024 arXiv

  89. [97]

    Wu, and Álvaro Tejero-Cantero

    Hanqi Zhou, Robert Bamler, Charley M. Wu, and Álvaro Tejero-Cantero. 2024. Predictive, scalable and interpretable knowledge tracing on structured domains. arXiv:2403.13179 [cs.LG] https://arxiv.org/abs/2403.13179

    2024 arXiv

  90. [98]

    Songlin Xu, Xinyu Zhang, and Lianhui Qin. 2024. EduAgent: Generative Student Agents in Learning. arXiv preprint arXiv:2404.07963 (2024)

    2024 arXiv

  91. [99]

    Xiaohan Xu, Ming Li, Chongyang Tao, Tao Shen, Reynold Cheng, Jinyang Li, Can Xu, Dacheng Tao, and Tianyi Zhou. 2024. A Survey on Knowledge Distillation of Large Language Models.ArXiv abs/2402.13116 (2024). https://api.semanticscholar. org/CorpusID:267760021

    2024 arXiv

  92. [100]

    Hanqi Yan, Qinglin Zhu, Xinyu Wang, Lin Gui, and Yulan He. 2024. Mirror: A Multiple-perspective Self-Reflection Method for Knowledge-rich Reasoning. arXiv preprint arXiv:2402.14963 (2024)

    2024 arXiv

  93. [101]

    Jackie Yang, Yingtian Shi, Yuhan Zhang, Karina Li, Daniel Wan Rosli, Anisha Jain, Shuning Zhang, Tianshi Li, James A Landay, and Monica S Lam. 2024. ReactGenie: A Development Framework for Complex Multimodal Interactions Using Large Language Models. In Proceedings of the CHI C...

    2024

  94. [102]

    Yang Yu, Yingbo Zhou, Yaokang Zhu, Yutong Ye, Liangyu Chen, and Mingsong Chen. 2024. ECKT: Enhancing Code Knowledge Tracing via Large Language Models. In Proceedings of the Annual Meeting of the Cognitive Science Society , Vol. 46

    2024

  95. [103]

    Michael V Yudelson, Kenneth R Koedinger, and Geoffrey J Gordon. 2013. Individ- ualized bayesian knowledge tracing models. In Artificial Intelligence in Education: 16th International Conference, AIED 2013, Memphis, TN, USA, July 9-13, 2013. Proceedings 16. Springer, 171–180

    2013

  96. [104]

    Murong Yue, Wijdane Mifdal, Yixuan Zhang, Jennifer Suh, and Ziyu Yao. 2024. MathVC: An LLM-Simulated Multi-Character Virtual Classroom for Mathematics Education. arXiv:2404.06711 [cs.CL] https://arxiv.org/abs/2404.06711 CHI ’25, April 26-May 1, 2025, Yokohama, Japan Xu, et al

    2024

  97. [106]

    Jiani Zhang, Xingjian Shi, Irwin King, and Dit-Yan Yeung. 2017. Dynamic Key- Value Memory Networks for Knowledge Tracing. arXiv:1611.08108 [cs.AI] https: //arxiv.org/abs/1611.08108

    2017 arXiv

  98. [107]

    Liang Zhang, Jionghao Lin, Conrad Borchers, John Sabatini, John Hollander, Meng Cao, and Xiangen Hu. 2024. Predicting Learning Performance with Large Language Models: A Study in Adult Literacy. In International Conference on Human-Computer Interaction. Springer, 333–353

    2024

  99. [110]

    Wazeer Deen Zulfikar, Samantha Chan, and Pattie Maes. 2024. Memoro: Using Large Language Models to Realize a Concise Interface for Real-Time Memory Augmentation. In Proceedings of the CHI Conference on Human Factors in Com- puting Systems. 1–18. A Appendix A.1 Deep Learning Ba...

    2024

  100. [113]

    Predict the student's performance: - Based on the analysis from steps 1 and 2, predict whether the student will answer each upcoming question correctly or incorrectly. CoT prompt LLM: Question 6: Correct, Reason: The student has previously answered questions related to the thr...

  101. [114]

    - Consider the complexity of the questions and the related lecture materials

    Analyze the student's past performance: - Identify the topics the student has performed well in and those they have struggled with. - Consider the complexity of the questions and the related lecture materials

  102. [115]

    - Identify if the upcoming questions are related to the topics the student has previously struggled with or excelled in

    Review the topics and related lecture materials of the upcoming questions: - Determine the difficulty level of the upcoming questions based on their topics and related lecture materials. - Identify if the upcoming questions are related to the topics the student has previously ...

  103. [116]

    Input" while the output of BertKT is denoted as

    Predict the student's performance: - Based on the analysis from steps 1 and 2, predict whether the student will answer each upcoming question correctly or incorrectly. CoT prompt with TIR LLM: Question 6: Correct, Reason: The student has previously answered questions related t...

    2025

  104. [2021]

    In Proceedings of the 29th ACM International Conference on Multimedia (Virtual Event, China) (MM ’21)

    Enhancing Knowledge Tracing via Adversarial Training. In Proceedings of the 29th ACM International Conference on Multimedia (Virtual Event, China) (MM ’21). Association for Computing Machinery, New York, NY, USA, 367–375. doi:10.1145/3474085.3475554

  105. [2023]

    Authorea Preprints (2023)

    A survey on large language models: Applications, challenges, limitations, and practical usage. Authorea Preprints (2023)

    2023

  106. [2024]

    In Companion Proceed- ings of the ACM on Web Conference 2024

    Large Language Model Powered Agents in the Web. In Companion Proceed- ings of the ACM on Web Conference 2024 . 1242–1245. CHI ’25, April 26-May 1, 2025, Yokohama, Japan Xu, et al

    2024

Pith tools

Reviewed August 9, 2026 · model on record in the stance chip above.