Pith. sign in

REVIEW 4 major objections 4 minor 2 cited by

Agent4Edu: Generating Learner Response Data by Generative Agents for Intelligent Education Systems

T0 review · 4 major / 4 minor · reviewed 2026-08-10 · deepseek-v4-flash

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

Pith's one-line read Agent4Edu claims that a single large-language-model agent can simulate a learner's choices, understanding, and answers well enough to improve adaptive testing.

desk verdict The framework is a plausible engineering contribution, but the main simulation evaluation likely leaks test answers through corrective reflection, so the held-out claim does not hold. read the letter →

arxiv 2501.10332 v2 pith:6ZRGGYNS submitted 2025-01-17 cs.CY cs.AI

Weibo Gao , Qi Liu , Linan Yue , Fangzhou Yao , Rui Lv , Zheng Zhang , Hao Wang , Zhenya Huang This is my paper · ORCID
classification cs.CYcs.AI
keywords learnerresponsesimulationgenerativeagentslargelanguagemodelspersonalizedlearningcomputerizedadaptivetestingprofileknowledgeproficiencyintelligenttutoringsystems
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

Agent4Edu is a simulator that turns a large language model into a stand-in for a specific learner, with the goal of generating future practice responses and making the learner's problem-solving process visible. The paper claims that this agent, initialized from real response records through profile, memory, and action modules, predicts future correctness better than supervised learner simulators and reproduces human-like exercise understanding. The authors position this as a way to close the gap between offline evaluation and online performance: personalized algorithms can be tested on configurable simulated learners, and simulated data can be added to real training data to improve computerized adaptive testing. If the claim holds, intelligent education systems could develop and validate new services without waiting for large amounts of fresh student practice data.

What carries the argument

The machinery is the three-module generative agent, one per learner. The learner profile module turns recorded practice logs into explicit tiers of activity, diversity, success rate, and preference, plus an IRT-estimated ability. The memory module stores recent and reinforced practice facts, updates a proficiency estimate with a cognitive diagnosis model, writes high-level summaries through reflection, and applies a forgetting function inspired by human memory decay. The action module then uses chain-of-thought reasoning to decide whether to attempt a recommended exercise, identify the knowledge concept being tested, and produce a solution with a predicted correctness label. Together the modules convert static response logs into a decision process that can be replayed against any personalized learning algorithm through the environment interface.

What would settle it

Retrain the adaptive-testing model on the original training data augmented with 20 responses per test learner drawn from that learner's historical success-rate distribution, with no agent reasoning, and compare the F1 changes with Table 4; if the gains persist, the reported improvement does not depend on the simulation machinery.

Watch

Extended reading notes

Core claim

The central discovery, stated on the paper's own terms, is that a learner's response behavior can be captured by a generative agent whose learner profile stores practice style and estimated ability, whose memory keeps reinforced facts, summaries, and evolving knowledge proficiency, and whose action module reads, understands, solves, and decides whether to attempt an exercise. In experiments, the agent with GPT-3.5-turbo reached 66.70% accuracy and 79.84% F1 on future-response prediction, ahead of the supervised baselines, and correctly identified the knowledge concept of an exercise in 73.88% of cases. The paper also reports that augmenting real training data with 20 agent-generated responses per test learner improved IRT-based adaptive testing F1 scores, with the largest gains at 2.39 percentage points for KLI at testing length five and 2.19 at length ten. These results are offered as evidence that the agent simulates not only correctness labels but also the understanding and analysis underlying human responses.

Load-bearing premise

The adaptive-testing improvement result assumes that responses generated by an agent initialized from a learner's own training records provide legitimate new evidence about that same learner, rather than restating the aggregate statistics already present in the training set.

Editorial extensions

If this is right

  • Personalized learning algorithms can be evaluated on simulated learners before deployment, reducing reliance on fresh real-world practice data.
  • Cold-start learners with no response history become simulable, because the agent can be initialized with a randomly generated profile.
  • Each simulated response carries an interpretable trace of exercise understanding and solution reasoning, not just a correctness label.
  • Merging agent-generated responses with real training data is reported to improve IRT-based computerized adaptive testing on the tested dataset.

Reading between the lines

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

  • Beyond the paper: the adaptive-testing improvements should be compared against a base-rate control, such as synthetic responses drawn from each learner's historical success rate, to confirm the agent adds learner-specific information rather than echoing training statistics.
  • Beyond the paper: the same open environment interface could test exercise recommendation and learning-path planning, which would show whether the simulated reasoning transfers beyond ability estimation.
  • Beyond the paper: the human-likeness evidence rests on LLM judges; human raters or an analysis of whether the agent's errors match the learner's error categories would be a stronger test.
  • Beyond the paper: a multi-agent classroom of interacting learners could model peer influence and competition, a direction the paper names as future work but does not evaluate.
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

4 major / 4 minor

Summary. The paper introduces Agent4Edu, an LLM-powered generative-agent framework for simulating learner response data in intelligent education systems. Each learner is represented by an agent with profile, memory, and action modules, initialized from the learner's real response records, and the agent can interact with personalized learning algorithms such as computerized adaptive testing. The authors report experiments on a proprietary dataset (EduData) in which Agent4Edu predicts held-out responses, performs zero-shot simulation, evaluates CAT strategies, and augments CAT training data. The main claims are that Agent4Edu outperforms existing learner simulation methods, produces human-like practice behavior, and can improve personalized learning algorithms through simulated data.

Significance. If the claims are substantiated, the paper would be a useful contribution: a generative-agent simulator for education could support training and evaluation of personalized learning algorithms without additional data collection, and the authors have made code, data, and an appendix publicly available. The design is thoughtful in combining psychological memory mechanisms, IRT-based ability estimates, and LLM-based reasoning into a single agent architecture. However, the empirical evidence as reported has several load-bearing ambiguities and contradictions that need to be resolved before the central claims can be accepted.

major comments (4)
  1. [§3.2 and §4.1] The evaluation may leak test-set labels through corrective reflection. In §3.2, the Action Module states that corrective reflection is triggered whenever the agent's predicted response does not match the real learner's response. In §4.1, test exercises are sent sequentially to the agent, but the paper never states that corrective reflection is disabled during this phase. The zero-shot section explicitly disables corrective reflection 'due to the absence of learner response data', implying that the default configuration includes it. If corrective reflection is active during the learner simulation evaluation, Table 1 does not measure held-out prediction but rather test-set adaptation. Please state explicitly whether corrective reflection and real-response-based memory updates are disabled during the learner simulation evaluation, and if they are not disabled, rerun the evaluation with them disabled.
  2. [§4.2 and Table 4] The CAT improvement experiment is ambiguous and potentially circular. For each test learner, the paper generates 20 simulated responses from that same learner's profile and merges them into EduData+. If the IRT model is retrained on EduData+ and then evaluated on the same test learners' real responses, the gains in Table 4 could reflect the model fitting each learner's own statistics rather than genuine transferable simulation signal. In addition, the caption states 'F1-score+ represents scores on EduData+', which suggests the improved metric is computed on the augmented training set rather than on held-out test records. Please clarify the evaluation protocol, and add a control that generates synthetic data only for learners in the training split and then evaluates on held-out test learners.
  3. [Table 1] The central 'outperforming existing learner simulation methods' claim rests on small margins (e.g., ACC 66.70 vs 65.72 for EERNN and 65.63 for DAISIM), but no variance or significance information is reported for these differences. The text says each baseline was run five times and averages were reported, but it does not state whether Agent4Edu itself was run multiple times. Please report standard deviations, confidence intervals, or significance tests for all rows, and state the number of runs for the agent.
  4. [§4.1, Zero-shot Simulation and Figure 2(b)] The zero-shot results as plotted contradict the claim of human-like indistinguishability. For the answering task, the agent's records were judged non-human (Lose) 42.67% of the time versus human (Win) 12.33%, and for summarization the corresponding values are Lose 37.33% versus Win 23.00%. The text concludes that the summarization is 'closely aligned with the real human responses', but the reported distribution shows the opposite tendency. The figure legend should clarify what the bars represent, and the claim about human-like zero-shot simulation should be substantially softened or replaced with a quantitative characterization of the failure modes.
minor comments (4)
  1. [Figure 2(c)] The ablation figure contains no numerical values and the x-axis labels ('complete w/o prof', etc.) are hard to read; please add a table with the exact accuracy values for each ablation condition.
  2. [§4.1, ROUGE-3] ROUGE-3 is described only as a measure of 'similarity between the simulated and real data distributions'; please specify exactly how ROUGE-3 is computed on response sequences, since ROUGE is normally used for text overlap.
  3. [Table 3] The counts in Table 3 are raw counts over agents, but the number of agents per cell and any measure of variability are not reported; please report proportions and sample sizes.
  4. [References] Several reference entries appear incomplete or duplicated (e.g., Atkinson 1968a/b and the two Wang et al. 2023 entries with nearly identical titles); please verify the bibliography against the cited sources.

Circularity Check

2 steps flagged · score 6.0 of 10

Test-time corrective reflection appears to use the real learner's label to correct the agent during the held-out simulation evaluation, so the headline ACC/F1 gains may reflect test-set adaptation rather than held-out prediction.

  1. fitted input called prediction [Section 3.2 (Action Module, Corrective Reflection) and Section 4.1 (Learner Simulation Evaluation, Table 1)]
    "Afterwards, the agent predicts whether its answer is correct (i.e., performance prediction). If the predicted response does not match the real learner’s response, a corrective reflection is triggered. ... For our generative agent, exercises from the test data are sequentially sent to it, and it performs the designed three actions to solve them."

    The evaluation treats the test exercises as unseen, but the agent's action module is defined to feed the real learner's response into a corrective reflection whenever the agent's prediction mismatches, and the test exercises are sent sequentially. The paper does not state that corrective reflection is disabled in this phase; it only says it is disabled in zero-shot simulation 'due to the absence of learner response data'. This implies the default configuration keeps corrective reflection active. Therefore the agent updates on the very test labels it is supposed to predict, so the ACC/F1 numbers in Table 1 are not pure held-out predictions but in-sample adaptation.

  2. fitted input called prediction [Section 4.2, Personalized Learning Algorithm Improvement (Table 4)]
    "Furthermore, for each learner in the test data, we simulate their responses to 20 randomly selected unseen exercises based on their profiles. Using this strategy, we generate simulated learner data, which are then merged with the training data from the original EduData to form the augmented dataset, EduData+."

    The agent profiles are initialized from each test learner's own response records, as described in Section 3.2. The 20 simulated responses added to the training set are therefore generated from the very learners on whom the CAT model is later evaluated. Retraining the IRT model on EduData+ injects learner-specific synthetic evidence for the test population, so the F1 gains in Table 4 can reflect the model echoing each test learner's own training statistics rather than learning from a general-purpose simulator. The experiment is presented as an improvement from 'high-quality learner response data', but the simulated data are derived from the target learners' own past responses, making the evaluation partially circular.

full rationale

Most of Agent4Edu's architecture is not circular: the profile, memory, and action modules form a genuine generative-agent design, and the paper evaluates against real human labels. However, a load-bearing part of the evaluation is. Section 3.2 defines Corrective Reflection as firing whenever the agent's predicted response mismatches the real learner's response, and Section 4.1 sends test exercises sequentially without stating that this reflection is disabled; the zero-shot section explicitly disables it only because no learner response data exist, implying default operation includes it. Thus, during the supposedly held-out simulation evaluation, the agent can be updated using test labels, contaminating the ACC/F1 improvements in Table 1 and the claim of outperforming existing learner simulation methods. The CAT improvement experiment in Section 4.2 has a related self-echoing problem: simulated responses are generated from each test learner's own profile and then added to the training set, so improved IRT predictions on those same learners may partly reflect the model being trained on learner-specific synthetic data rather than independent simulation signal. Self-citations to IRT and DNeuralCDM are used as profiling tools, not as proof of the main result, so they are not load-bearing. The overall circularity score is 6 because one or more of the headline 'predictions' reduce to test-set adaptation or self-echoing of learner statistics by construction, while the framework itself retains substantial independent content.

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

The central claim rests on a few hand-chosen memory thresholds, an assumed forgetting curve, and the premise that LLM behavior generalizes to unseen exercises. No new entities are introduced; the system composes existing models.

free parameters (5)
  • Short-term memory size = s=5
    Set empirically; no sensitivity analysis reported.
  • Memory enhancement threshold = F=5
    Reinforced fact threshold, set empirically.
  • Forgetting threshold = lambda=0.99
    Threshold in the forgetting function g(i)=1/(1+e^-(n-i)); set empirically.
  • Number of simulated responses per test learner = 20
    Used in the CAT improvement experiment; arbitrary choice.
  • LLM temperature = 0 for agent, 0.5 for annotators
    Temperature set to 0 to reduce randomness; annotators use 0.5.
assumptions (5)
  • domain assumption Each learner submits each exercise at most once, so repeated submissions are not modeled.
    Stated in Section 3.1; simplifies the task but ignores a real behavior.
  • domain assumption The LLM can stand in for a human learner's reading, analysis, and solution process.
    Core premise of the agent design; not independently validated.
  • ad hoc to paper The forgetting curve g(l)=1/(1+e^-(n-i)) captures human memory decay.
    Chosen functional form with fitted threshold lambda; not derived from cognitive data.
  • domain assumption IRT and DNeuralCDM provide valid estimates of ability and knowledge proficiency.
    These are standard models used as tools, trained on the same response data; their validity is assumed.
  • domain assumption GPT-3.5-turbo as annotator is a valid judge of whether responses are human-written.
    Used in zero-shot evaluation without calibration against human raters.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Agent4Edu: Generating Learner Response Data by Generative Agents for Intelligent Education Systems." pith.science (2026). https://pith.science/paper/6ZRGGYNS

@misc{pith2026250110332,
  author       = {Pith},
  title        = {Pith review of: Agent4Edu: Generating Learner Response Data by Generative Agents for Intelligent Education Systems},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/6ZRGGYNS}},
  note         = {Machine review of arXiv:2501.10332}
}
read the original abstract

Personalized learning represents a promising educational strategy within intelligent educational systems, aiming to enhance learners' practice efficiency. However, the discrepancy between offline metrics and online performance significantly impedes their progress. To address this challenge, we introduce Agent4Edu, a novel personalized learning simulator leveraging recent advancements in human intelligence through large language models (LLMs). Agent4Edu features LLM-powered generative agents equipped with learner profile, memory, and action modules tailored to personalized learning algorithms. The learner profiles are initialized using real-world response data, capturing practice styles and cognitive factors. Inspired by human psychology theory, the memory module records practice facts and high-level summaries, integrating reflection mechanisms. The action module supports various behaviors, including exercise understanding, analysis, and response generation. Each agent can interact with personalized learning algorithms, such as computerized adaptive testing, enabling a multifaceted evaluation and enhancement of customized services. Through a comprehensive assessment, we explore the strengths and weaknesses of Agent4Edu, emphasizing the consistency and discrepancies in responses between agents and human learners. The code, data, and appendix are publicly available at https://github.com/bigdata-ustc/Agent4Edu.

Figures

Figures reproduced from arXiv: 2501.10332 by the authors.

Figure 1
Figure 1. All the prompts are listed in Appendix C. [PITH_FULL_IMAGE:figures/full_fig_p003_1.png] view at source ↗
Figure 1
Figure 1. The overall framework of Agent4Edu. tailed exposition, refer to Appendix A.1. Additionally, to en￾sure broad applicability and protect privacy, certain personal identifiers (such as name, gender, age, and occupation) are intentionally anonymized in this work (Zhang et al. 2023a; Li et al. 2023). While these attributes may help shape other types of agents, they are not primary factors affecting prac￾tice performance … view at source ↗
Figure 2
Figure 2. (a) Comparison between the success rate distributions of ground-truth and agent-simulated response data. (b) Using [PITH_FULL_IMAGE:figures/full_fig_p007_2.png] view at source ↗

Discussion (0). Continue with ORCID to comment.

Forward citations

Cited by 2 Pith papers

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

  1. CoderAgent: Simulating Student Behavior for Personalized Programming Learning with Large Language Models

    cs.AI 2025-05 conditional novelty 6.0 of 10

    A new LLM-based agent, CoderAgent, simulates students' iterative programming process (why, how, where, what to modify) and outperforms baselines on predicting next code edits, though gains are modest.

  2. AgentSME for Simulating Diverse Communication Modes in Smart Education

    cs.AI 2025-08 conditional novelty 4.0 of 10

    A one-round peer-exchange protocol improves accuracy of six LLMs on a Chinese sociology multiple-choice benchmark, and DeepSeek shows the largest lexical diversity.

Reference graph

Works this paper leans on

53 extracted references · 33 canonical work pages · cited by 2 Pith papers

  1. [1]

    , " * write output.state after.block = add.period write newline

    ENTRY address archivePrefix author booktitle chapter edition editor eid eprint howpublished institution isbn journal key month note number organization pages publisher school series title type volume year label extra.label sort.label short.list INTEGERS output.state before.all mid.sentence after.sentence after.block FUNCTION init.state.consts #0 'before.a...

  2. [2]

    write newline

    " write newline "" before.all 'output.state := FUNCTION n.dashify 't := "" t empty not t #1 #1 substring "-" = t #1 #2 substring "--" = not "--" * t #2 global.max substring 't := t #1 #1 substring "-" = "-" * t #2 global.max substring 't := while if t #1 #1 substring * t #2 global.max substring 't := if while FUNCTION word.in bbl.in capitalize " " * FUNCT...

  3. [3]

    Atkinson, R. C. 1968 a . Human memory: A proposed system and its control processes. The Psychology of Learning and Motivation, 2

    work page 1968

  4. [4]

    Atkinson, R. C. 1968 b . A proposed system and its control processes. The Psychology of Learning and Motivation, 2

    work page 1968

  5. [5]

    Averell, L.; and Heathcote, A. 2011. The form of the forgetting curve and the fate of memories. Journal of mathematical psychology, 55(1): 25--35

    work page 2011

  6. [6]

    Baidoo-Anu, D.; and Ansah, L. O. 2023. Education in the era of generative artificial intelligence (AI): Understanding the potential benefits of ChatGPT in promoting teaching and learning. Journal of AI, 7(1): 52--62

    work page 2023

  7. [7]

    Baker, F. B. 2001. The basics of item response theory. ERIC

    work page 2001

  8. [8]

    Bi, H.; Ma, H.; Huang, Z.; Yin, Y.; Liu, Q.; Chen, E.; Su, Y.; and Wang, S. 2020. Quality meets diversity: A model-agnostic framework for computerized adaptive testing. In 2020 IEEE International Conference on Data Mining (ICDM), 42--51. IEEE

    work page 2020

Show all 53 references
  1. [9]

    D.; Dhariwal, P.; Neelakantan, A.; Shyam, P.; Sastry, G.; Askell, A.; et al

    Brown, T.; Mann, B.; Ryder, N.; Subbiah, M.; Kaplan, J. D.; Dhariwal, P.; Neelakantan, A.; Shyam, P.; Sastry, G.; Askell, A.; et al. 2020. Language models are few-shot learners. Advances in neural information processing systems, 33: 1877--1901

    2020

  2. [10]

    Chang, H.-H.; and Ying, Z. 1996. A global information approach to computerized adaptive testing. Applied Psychological Measurement, 20(3): 213--229

    1996

  3. [11]

    Chen, X.; Wu, L.; Liu, F.; Chen, L.; Zhang, K.; Hong, R.; and Wang, M. 2024. Disentangling Cognitive Diagnosis with Limited Exercise Labels. Advances in Neural Information Processing Systems, 36

    2024

  4. [12]

    Cheng, K.; Peng, L.; Wang, P.; Ye, J.; Sun, L.; and Du, B. 2024. DyGKT: Dynamic Graph Learning for Knowledge Tracing. In Proceedings of the 30th ACM SIGKDD Conference on Knowledge Discovery and Data Mining, 409--420

    2024

  5. [13]

    Cowan, N. 2008. What are the differences between long-term, short-term, and working memory? Progress in brain research, 169: 323--338

    2008

  6. [14]

    Dan, Y.; Lei, Z.; Gu, Y.; Li, Y.; Yin, J.; Lin, J.; Ye, L.; Tie, Z.; Zhou, Y.; Wang, Y.; et al. 2023. Educhat: A large-scale language model-based chatbot system for intelligent education. arXiv preprint arXiv:2308.02773

    2023 arXiv

  7. [15]

    Gao, C.; Lan, X.; Lu, Z.; Mao, J.; Piao, J.; Wang, H.; Jin, D.; and Li, Y. 2023. S ^ 3 : Social-network Simulation System with Large Language Model-Empowered Agents. arXiv preprint arXiv:2307.14984

    2023 arXiv

  8. [16]

    Gao, W.; Liu, Q.; Huang, Z.; Yin, Y.; Bi, H.; Wang, M.-C.; Ma, J.; Wang, S.; and Su, Y. 2021. RCD: Relation map driven cognitive diagnosis for intelligent education systems. In Proceedings of the 44th international ACM SIGIR conference on research and development in informatio...

    2021

  9. [17]

    Huang, X.; Lian, J.; Lei, Y.; Yao, J.; Lian, D.; and Xie, X. 2023. Recommender ai agent: Integrating large language models for interactive recommendations. arXiv preprint arXiv:2308.16505

    2023 arXiv

  10. [18]

    Huang, Z.; Liu, Q.; Chen, Y.; Wu, L.; Xiao, K.; Chen, E.; Ma, H.; and Hu, G. 2020. Learning or forgetting? a dynamic approach for tracking the knowledge proficiency of students. ACM Transactions on Information Systems (TOIS), 38(2): 1--33

    2020

  11. [19]

    Huang, Z.; Liu, Q.; Zhai, C.; Yin, Y.; Chen, E.; Gao, W.; and Hu, G. 2019. Exploring multi-objective exercise recommendations in online education systems. In Proceedings of the 28th ACM International Conference on Information and Knowledge Management, 1261--1270

    2019

  12. [20]

    Jin, D.; Mehri, S.; Hazarika, D.; Padmakumar, A.; Lee, S.; Liu, Y.; and Namazifar, M. 2023. Data-efficient alignment of large language models with human feedback through natural language. arXiv preprint arXiv:2311.14543

    2023 arXiv

  13. [21]

    Kieser, F.; Wulff, P.; Kuhn, J.; and K \"u chemann, S. 2023. Educational data augmentation in physics education research using ChatGPT. Physical Review Physics Education Research, 19(2): 020150

    2023

  14. [22]

    Li, H.; Xu, T.; Zhang, C.; Chen, E.; Liang, J.; Fan, X.; Li, H.; Tang, J.; and Wen, Q. 2024. Bringing generative AI to adaptive learning in education. arXiv preprint arXiv:2402.14601

    2024 arXiv

  15. [23]

    Li, Y.; Chen, X.; Zhao, H.; Gong, J.; Zhou, G.; Rossano, F.; and Zhu, Y. 2023. Understanding Embodied Reference with Touch-Line Transformer. In ICLR

    2023

  16. [24]

    ???? SocraticLM: Exploring Socratic Personalized Teaching with Large Language Models

    Liu, J.; Huang, Z.; Xiao, T.; Sha, J.; Wu, J.; Liu, Q.; Wang, S.; and Chen, E. ???? SocraticLM: Exploring Socratic Personalized Teaching with Large Language Models. In The Thirty-eighth Annual Conference on Neural Information Processing Systems

  17. [25]

    Liu, Q.; Tong, S.; Liu, C.; Zhao, H.; Chen, E.; Ma, H.; and Wang, S. 2019. Exploiting cognitive structure for adaptive learning. In Proceedings of the 25th ACM SIGKDD international conference on knowledge discovery & data mining, 627--635

    2019

  18. [26]

    M.; Yang, D.; and Vosoughi, S

    Liu, R.; Yang, R.; Jia, C.; Zhang, G.; Zhou, D.; Dai, A. M.; Yang, D.; and Vosoughi, S. 2023. Training socially aligned language models in simulated human society. arXiv preprint arXiv:2305.16960

    2023 arXiv

  19. [27]

    Long, X.; Zeng, J.; Meng, F.; Ma, Z.; Zhang, K.; Zhou, B.; and Zhou, J. 2024. Generative multi-modal knowledge retrieval with large language models. In Proceedings of the AAAI Conference on Artificial Intelligence, volume 38, 18733--18741

    2024

  20. [28]

    Lord, F. M. 2012. Applications of item response theory to practical testing problems. Routledge

    2012

  21. [29]

    K.; Parodi, F.; Liu, T.; Lange, R

    Matelsky, J. K.; Parodi, F.; Liu, T.; Lange, R. D.; and Kording, K. P. 2023. A large language model-assisted education tool to provide feedback on open-ended responses. arXiv preprint arXiv:2308.02439

    2023 arXiv

  22. [30]

    Ouyang, L.; Wu, J.; Jiang, X.; Almeida, D.; Wainwright, C.; Mishkin, P.; Zhang, C.; Agarwal, S.; Slama, K.; Ray, A.; et al. 2022. Training language models to follow instructions with human feedback. Advances in neural information processing systems, 35: 27730--27744

    2022

  23. [31]

    Pandey, S.; and Karypis, G. 2019. A self-attentive model for knowledge tracing. In 12th International Conference on Educational Data Mining, EDM 2019, 384--389. International Educational Data Mining Society

    2019

  24. [32]

    S.; O'Brien, J.; Cai, C

    Park, J. S.; O'Brien, J.; Cai, C. J.; Morris, M. R.; Liang, P.; and Bernstein, M. S. 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

  25. [33]

    J.; and Sohl-Dickstein, J

    Piech, C.; Bassen, J.; Huang, J.; Ganguli, S.; Sahami, M.; Guibas, L. J.; and Sohl-Dickstein, J. 2015. Deep knowledge tracing. Advances in neural information processing systems, 28

    2015

  26. [34]

    Qadir, J. 2023. Engineering education in the era of ChatGPT: Promise and pitfalls of generative AI for education. In 2023 IEEE Global Engineering Education Conference (EDUCON), 1--9. IEEE

    2023

  27. [35]

    Qian, C.; Cong, X.; Yang, C.; Chen, W.; Su, Y.; Xu, J.; Liu, Z.; and Sun, M. 2023. Communicative agents for software development. arXiv preprint arXiv:2307.07924

    2023 arXiv

  28. [36]

    M.; and Watanobe, Y

    Rahman, M. M.; and Watanobe, Y. 2023. ChatGPT for education and research: Opportunities, threats, and strategies. Applied Sciences, 13(9): 5783

    2023

  29. [37]

    Reddy, S.; Levine, S.; and Dragan, A. 2017. Accelerating human learning with deep reinforcement learning. In NIPS workshop: teaching machines, robots, and humans

    2017

  30. [38]

    Su, Y.; Liu, Q.; Liu, Q.; Huang, Z.; Yin, Y.; Chen, E.; Ding, C.; Wei, S.; and Hu, G. 2018. Exercise-enhanced sequential modeling for student performance prediction. In Proceedings of the AAAI conference on artificial intelligence, volume 32

    2018

  31. [39]

    Wang, F.; Gao, W.; Liu, Q.; Li, J.; Zhao, G.; Zhang, Z.; Huang, Z.; Zhu, M.; Wang, S.; Tong, W.; et al. 2024 a . A Survey of Models for Cognitive Diagnosis: New Developments and Future Directions. arXiv preprint arXiv:2407.05458

    2024 arXiv

  32. [40]

    Wang, F.; Huang, Z.; Liu, Q.; Chen, E.; Yin, Y.; Ma, J.; and Wang, S. 2023 a . Dynamic cognitive diagnosis: An educational priors-enhanced deep knowledge tracing perspective. IEEE Transactions on Learning Technologies, 16(3): 306--323

    2023

  33. [41]

    Wang, G.; Xie, Y.; Jiang, Y.; Mandlekar, A.; Xiao, C.; Zhu, Y.; Fan, L.; and Anandkumar, A. 2023 b . Voyager: An open-ended embodied agent with large language models. arXiv preprint arXiv:2305.16291

    2023 arXiv

  34. [42]

    Wang, L.; Ma, C.; Feng, X.; Zhang, Z.; Yang, H.; Zhang, J.; Chen, Z.; Tang, J.; Chen, X.; Lin, Y.; et al. 2024 b . A survey on large language model based autonomous agents. Frontiers of Computer Science, 18(6): 1--26

    2024

  35. [44]

    X.; et al

    Wang, L.; Zhang, J.; Yang, H.; Chen, Z.; Tang, J.; Zhang, Z.; Chen, X.; Lin, Y.; Song, R.; Zhao, W. X.; et al. 2023 d . User behavior simulation with large language model based agents. arXiv preprint arXiv:2306.02552

    2023 arXiv

  36. [45]

    V.; Zhou, D.; et al

    Wei, J.; Wang, X.; Schuurmans, D.; Bosma, M.; Xia, F.; Chi, E.; Le, Q. V.; Zhou, D.; et al. 2022. Chain-of-thought prompting elicits reasoning in large language models. Advances in neural information processing systems, 35: 24824--24837

    2022

  37. [46]

    Wu, Q.; Bansal, G.; Zhang, J.; Wu, Y.; Zhang, S.; Zhu, E.; Li, B.; Jiang, L.; Zhang, X.; and Wang, C. 2023. Autogen: Enabling next-gen llm applications via multi-agent conversation framework. arXiv preprint arXiv:2308.08155

    2023 arXiv

  38. [47]

    Xu, S.; Zhang, X.; and Qin, L. 2024. EduAgent: Generative Student Agents in Learning. arXiv preprint arXiv:2404.07963

    2024 arXiv

  39. [48]

    Yao, F.; Liu, Q.; Yue, L.; Gao, W.; Li, J.; Li, X.; and He, Y. 2024. Adard: An adaptive response denoising framework for robust learner modeling. In Proceedings of the 30th ACM SIGKDD Conference on Knowledge Discovery and Data Mining, 3886--3895

    2024

  40. [49]

    Yue, L.; Liu, Q.; Du, Y.; Gao, W.; Liu, Y.; and Yao, F. 2023. Fedjudge: Federated legal large language model. arXiv preprint arXiv:2309.08173

    2023 arXiv

  41. [50]

    Zhang, A.; Sheng, L.; Chen, Y.; Li, H.; Deng, Y.; Wang, X.; and Chua, T.-S. 2023 a . On generative agents in recommendation. arXiv preprint arXiv:2310.10108

    2023 arXiv

  42. [51]

    X.; Lin, L.; and Wen, J.-R

    Zhang, J.; Hou, Y.; Xie, R.; Sun, W.; McAuley, J.; Zhao, W. X.; Lin, L.; and Wen, J.-R. 2023 b . Agentcf: Collaborative learning with autonomous language agents for recommender systems. arXiv preprint arXiv:2310.09233

    2023 arXiv

  43. [52]

    Zhang, J.; Shi, X.; King, I.; and Yeung, D.-Y. 2017. Dynamic key-value memory networks for knowledge tracing. In Proceedings of the 26th international conference on World Wide Web, 765--774

    2017

  44. [53]

    Zhao, G.; Huang, Z.; Zhuang, Y.; Liu, J.; Liu, Q.; Liu, Z.; Wu, J.; and Chen, E. 2023. Simulating Student Interactions with Two-stage Imitation Learning for Intelligent Educational Systems. In Proceedings of the 32nd ACM International Conference on Information and Knowledge Ma...

    2023

  45. [54]

    Zhuang, Y.; Liu, Q.; Zhao, G.; Huang, Z.; Huang, W.; Pardos, Z.; Chen, E.; Wu, J.; and Li, X. 2024. A Bounded Ability Estimation for Computerized Adaptive Testing. Advances in Neural Information Processing Systems, 36

    2024

Pith tools

Reviewed August 10, 2026 · model on record in the stance chip above.