Pith. sign in

REVIEW 3 major objections 3 minor 51 references

Machine Assistant with Reliable Knowledge: Enhancing Student Learning via RAG-based Retrieval

T0 review · 3 major / 3 minor · reviewed 2026-08-06 · deepseek-v4-flash

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

Pith's one-line read This paper introduces MARK, a retrieval-augmented chatbot that grounds answers in a curated course knowledge base, and claims this design reduces hallucination risk and can substitute for office hours and routine technical support.

desk verdict A clear system description of a standard hybrid RAG chatbot, but the central reliability claims rest entirely on anecdote and are not measured. read the letter →

arxiv 2506.23026 v1 pith:VPNLVFDT submitted 2025-06-28 cs.IR

Yongsheng Lian This is my paper
classification cs.IR
keywords retrieval-augmentedgenerationRAGhallucinationreductionhybridsearchBM25denseretrievaleducationalchatbotfeedbackloop
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

MARK is a retrieval-augmented chatbot that answers each student or customer question by hunting through a curated knowledge base with both exact keyword matching and semantic vector search, merging the hits, and handing the top passages to a large language model as the basis for a concise answer. The paper claims this grounded design produces more accurate and trustworthy responses and reduces the hallucination risk of answering from memory alone. A feedback loop lets instructors rewrite poor answers, and the corrections are indexed back into the corpus, so the system improves with use. Deployed in a Fluid Mechanics course as a stand-in for office hours and in a customer-support setting, MARK reportedly handled a broad range of routine queries. A sympathetic reader would take the contribution as a recipe: RAG plus instructor curation and feedback can give students a reliable, scalable 24/7 assistant.

What carries the argument

The load-bearing mechanism is a hybrid retrieval corpus with a feedback loop. Queries run in parallel through BM25, which matches exact terms, and a dense vector search over 512-dimensional Jina embeddings indexed with approximate-nearest-neighbor search; the top results are merged, re-ranked, and the top k passages become context for the language model. The prompt asks the model to be a patient teaching assistant, refer to prior conversation, politely decline out-of-knowledge questions, and be concise. Instructor corrections entered through a monitoring panel are appended to the raw document store, and both the sparse and dense indexes are regenerated, making the knowledge base the system's memory rather than the model's parameters.

What would settle it

Pose a set of course questions whose correct answers are deliberately absent from the knowledge base and a matched set whose answers are present, then check whether the system frequently answers the absent set with confident but wrong content instead of politely declining; any substantial rate of unsupported answers would falsify the groundedness claim.

Watch

Extended reading notes

Core claim

The paper's central claim is that a retrieval-augmented generator grounded in a curated, instructor-maintained knowledge base can deliver accurate and trustworthy responses, sufficiently so that it can stand in for office hours and handle routine technical support. The mechanism is MARK: BM25 keyword retrieval and dense embedding retrieval run in parallel, the results are merged and re-ranked, the top passages go to a language model under a short prompt, and instructor corrections are written back into the corpus. The paper reports deployment in a Fluid Mechanics course and in a customer-support setting, where the system successfully addressed a broad range of student queries and answered routine customer questions. The intended consequence is reduced hallucination relative to a bare language model, because outputs are anchored to curated passages rather than the model's parametric memory.

Load-bearing premise

The entire reliability claim rests on two premises that the paper does not measure: that the language model, given the retrieved chunks and the MARK prompt, will stick to those chunks rather than its own memorized knowledge, and that the curated corpus contains accurate and complete answers; if either premise fails, the claimed reduction in hallucination is not assured.

Editorial extensions

If this is right

  • An instructor can deploy MARK as a 24/7 substitute for routine office hours, absorbing logistical and homework queries without additional teaching-staff time.
  • Every corrected answer becomes part of the retrieval corpus, so the system is expected to improve with use.
  • Hybrid BM25-plus-dense retrieval covers both exact-term queries and paraphrased semantic queries, making the assistant more robust to different question styles.
  • Swapping the curated corpus lets the same system serve technical support, including account lookups and troubleshooting, with natural-language answers.
  • MARK is not a full replacement for human support, because some students and customers still prefer or request interaction with a real person.

Reading between the lines

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

  • A testable implication left implicit: if the feedback loop is the main driver of reliability, then a version of MARK without instructor corrections should show measurably more hallucination on the same corpus; the paper does not isolate this effect.
  • The same design would plausibly transfer to other long-tail knowledge domains, such as internal policy help or clinical triage, where the corpus can be kept curated and current; that extension is not claimed by the paper.
  • The paper's own limitation note points to multimodal input as a natural next step, since text-only retrieval cannot catch student errors in equations or diagrams.
  • The 384-token chunk length and the practice of including section headings are testable design decisions; ablating them would reveal how much of the reliability comes from retrieval parameters rather than from the language model.
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 / 3 minor

Summary. The manuscript describes MARK, a retrieval-augmented generation (RAG) system that combines BM25 sparse retrieval with FAISS-based dense retrieval, a re-ranking stage, an instructor-feedback loop, and a web frontend. The authors report a pilot deployment in a Fluid Mechanics course and a technical-support use case, and they conclude that MARK 'is able to generate more accurate and trustworthy responses, reducing the risk of hallucination.' Sections 2 and 4 present standard retrieval and backend techniques; Sections 3 and 5 describe the interfaces and anecdotal deployments.

Significance. The paper's potential contribution is a deployable RAG system for education with instructor control and feedback. The system is publicly accessible, and the retrieval implementation follows well-established methods. However, the paper offers no quantitative evaluation of retrieval effectiveness, answer faithfulness, or hallucination rates, and the only evidence in Section 5 is narrative and illustrative. As a result, the central claims of accuracy, reliability, and reduced hallucination are not demonstrated; the significance of the work to the IR community is therefore not established.

major comments (3)
  1. [Section 5 (Results)] Section 5 contains no quantitative evaluation. The abstract and conclusion claim that MARK 'successfully addressed a broad range of student queries' and 'is able to generate more accurate and trustworthy responses, reducing the risk of hallucination,' yet Section 5.1 offers only a narrative description and Section 5.2 presents four illustrative dialogues. There are no query counts, accuracy or answer-faithfulness metrics, retrieval precision/recall numbers, hallucination-rate measurements, or comparisons against a no-RAG baseline or alternative retrievers. These claims therefore have no evidential support.
  2. [Sections 4.3 and 3.2] The prompt in Section 4.3 does not instruct the LLM to answer exclusively from retrieved content; it says only to 'politely decline if a question is outside your knowledge' and to 'be concise.' Moreover, Section 3.2 exposes a tunable parameter controlling how much external knowledge the LLM may use (0-100), but the deployed value is never stated. Thus the mechanism claimed to reduce hallucination is not guaranteed by the design, and no faithfulness measurement is reported to verify that responses are grounded in the retrieved passages.
  3. [Section 5.1] The Fluid Mechanics corpus was seeded with 'curated answers to student questions by instructors.' As a result, some successful responses may have been pre-written text retrieved from the corpus rather than answers generated by the LLM from retrieved evidence. The paper does not distinguish these cases, so the classroom results do not directly support the claim that RAG improves LLM generation. The manuscript's own statement that 'Overall adoption during the 2023 pilot phase was limited' further weakens the claim of successful deployment.
minor comments (3)
  1. [Eq. (6)] The BM25 IDF formula in Eq. (6) appears to include an extra '+1' term inside the logarithm; please check whether this is intentional or a typesetting error relative to the standard formulation.
  2. [Throughout] There are several typos: 'weigh' for 'weight', 'acorss' for 'across', 'docment' for 'document', 'conduced' for 'conducted', and 'BERTA' for 'BERT' in the example list of cross-encoder models in Eq. (12).
  3. [Sections 2 and 4.2] The paper does not report the values of hyperparameters such as k1, b, hybrid fusion weights, or top-k; please state them explicitly or clarify that system defaults were used.

Circularity Check

1 steps flagged · score 4.0 of 10

The pilot evaluation is partly self-referential: the Fluid Mechanics corpus is seeded with instructor-written answers to student questions, so the reported successful responses can be retrieved from the answer store rather than generated; no faithfulness or hallucination metrics separate the two.

  1. fitted input called prediction [Section 5.1 (MARK as Virtual Tutor) and Section 6 (Conclusion)]
    "We created the basic RAG database using class syllabus, class notes, and homework problems. The database is further enhanced with curated answers to student questions by instructors. ... These questions were typically brief, such as "How to solve Homework 5.6," reflecting a desire for direct assistance. MARK responded with complete solutions rather than offering step-by-step guidance. ... By integrating carefully curated knowledge with LLMs, MARK is able to generate more accurate and trustworthy responses, reducing the risk of hallucination."

    The evidence offered for the system's reliability is that MARK answered student questions, but the retrieval corpus explicitly contains instructor-written answers to student questions. For queries like 'How to solve Homework 5.6,' the top retrieved chunk can be the pre-seeded solution, so the 'successful' response is the stored answer replayed through the LLM rather than a grounded synthesis that demonstrates hallucination reduction. The conclusion then credits this curated-input replay as evidence that MARK generates 'more accurate and trustworthy responses.' Because no retrieval-precision, answer-faithfulness, or hallucination measurements are reported, the demonstrated success reduces, at least in part, to the corpus contents by construction.

full rationale

The paper is mostly a system description with standard components (BM25, FAISS, cross-encoder re-ranking, OpenAI API) and cites external, independent literature for the general claim that RAG reduces hallucination, so there is no self-citation chain or imported uniqueness theorem. The central architectural claim is not circular in the formal sense. However, the local evaluation in Section 5 is at least partly self-referential: the knowledge base is seeded with instructor-written answers to student questions, and the anecdotal successes reported there are not separated from simple retrieval of those seeded answers. Additional evidential gaps—the Section 4.3 prompt does not literally instruct the model to answer exclusively from retrieved content despite the surrounding claim, and Section 3.2 exposes an unstated 0-100 knob for how much external LLM knowledge is allowed—undermine confidence but are not themselves circularity. The paper also openly notes limited adoption ('Overall adoption during the 2023 pilot phase was limited') and that MARK 'cannot fully replace human support,' which is consistent with a modest, partially self-referential demonstration rather than a forced derivation. Overall circularity is moderate.

Assumptions & free parameters 6 free parameters · 5 assumptions · 0 invented entities

The system's core claims rest on standard IR mathematics plus unverified domain assumptions about corpus accuracy, prompt grounding, and feedback improvement. Several implementation parameters are chosen by hand or from unreported experiments, and no new physical or formal entities are introduced.

free parameters (6)
  • Maximum chunk length = 384 tokens
    Set in Section 4.1 based on unreported experiments; affects retrieval granularity and answer length.
  • Embedding dimension = 512
    Determined by jina-embeddings-v2-small-en (Section 4.1); claimed as a good balance without shown benchmarks.
  • External-knowledge usage slider = 0 to 100
    Instructor-controlled parameter in Section 3.2; its effect on answer reliability is untested.
  • BM25 k1 = 1.5 (default)
    Term-frequency saturation in Eq. 5; standard default, not tuned.
  • BM25 b = 0.75 (default)
    Length normalization in Eq. 5; standard default, not tuned.
  • Hybrid fusion weights and top-k = unspecified
    Section 4.2 merges BM25 and FAISS results but does not state the number of candidates or fusion weighting.
assumptions (5)
  • standard math TF-IDF, BM25, cosine similarity, and FAISS ANN are correct and appropriate retrieval tools.
    Section 2 reviews standard IR methods; the correctness of these methods is assumed from the literature.
  • domain assumption The curated knowledge base (syllabus, class notes, homework, instructor answers) is accurate and sufficient for answering student questions.
    Section 5.1 builds the RAG DB on these sources; the paper gives no verification of coverage or correctness.
  • domain assumption Retrieval of relevant chunks plus a one-sentence prompt makes the LLM ground its answer and suppress hallucination.
    Section 4.3 states this assumption; no faithfulness or hallucination evaluation is provided.
  • ad hoc to paper Instructor corrections inserted into the corpus improve future responses.
    Section 4.4 claims adaptive refinement; no data show that inserted corrections are retrieved and used correctly.
  • ad hoc to paper Including section headings in chunks and using 384-token chunks improves retrieval.
    Section 4.1 asserts this 'based on our experiments' without reporting the experiments.

how reviews work

0 comments
Cite this review

Pith. "Pith review of Machine Assistant with Reliable Knowledge: Enhancing Student Learning via RAG-based Retrieval." pith.science (2026). https://pith.science/paper/VPNLVFDT

@misc{pith2026250623026,
  author       = {Pith},
  title        = {Pith review of: Machine Assistant with Reliable Knowledge: Enhancing Student Learning via RAG-based Retrieval},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/VPNLVFDT}},
  note         = {Machine review of arXiv:2506.23026}
}
read the original abstract

We present Machine Assistant with Reliable Knowledge (MARK), a retrieval-augmented question-answering system designed to support student learning through accurate and contextually grounded responses. The system is built on a retrieval-augmented generation (RAG) framework, which integrates a curated knowledge base to ensure factual consistency. To enhance retrieval effectiveness across diverse question types, we implement a hybrid search strategy that combines dense vector similarity with sparse keyword-based retrieval. This dual-retrieval mechanism improves robustness for both general and domain-specific queries. The system includes a feedback loop in which students can rate responses and instructors can review and revise them. Instructor corrections are incorporated into the retrieval corpus, enabling adaptive refinement over time. The system was deployed in a classroom setting as a substitute for traditional office hours, where it successfully addressed a broad range of student queries. It was also used to provide technical support by integrating with a customer-specific knowledge base, demonstrating its ability to handle routine, context-sensitive tasks in applied domains. MARK is publicly accessible at https://app.eduquery.ai.

Figures

Figures reproduced from arXiv: 2506.23026 by the authors.

Figure 1
Figure 1. Interface of MARK with Text Entry and User Feedback [PITH_FULL_IMAGE:figures/full_fig_p007_1.png] view at source ↗
Figure 2
Figure 2. The four frontend modules of the instructor interface [PITH_FULL_IMAGE:figures/full_fig_p007_2.png] view at source ↗
Figure 3
Figure 3. Interface of Create a New Chatbot Chatbot Training Panel [PITH_FULL_IMAGE:figures/full_fig_p008_3.png] view at source ↗
Figures from the paper (3 more)
Figure 4
Figure 4. Figure 4: Interface of Uploading Contents for RAG Chatbot Testing Panel Following content ingestion, the chatbot can be evaluated through the Testing Panel, which is designed to simulate real-world user interactions. This interface mirrors the layout and functionality of the stu…
Figure 5
Figure 5. Figure 5: Interface of Monitoring AI Responses 4 Backend Development The backend architecture of MARK is implemented using the Python Flask web framework. Flask is a lightweight and efficient foundation for handling HTTP requests and serving the application’s API endpoints. This…
Figure 6
Figure 6. Figure 6: The four backend modules 4.1 RAG Database Preparation The uploaded documents from the frontend are processed through a backend API to prepare them for embedding and retrieval. For plain text and comma-separated values files, the content is segmented directly into small…

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

51 extracted references · 36 canonical work pages

  1. [1]

    C. Binkley. College students are still struggling with basic math. professors blame the pandemic. https: //apnews.com/article/college-math-test-help-6cca6a5e873d5aeb5e75b4f94125d48c , 2023. As- sociated Press, Accessed: 2025-04-29

    work page 2023

  2. [2]

    National Assessment of Educational Progress (NAEP)

    Naep mathematics: National average scores, 2022. National Assessment of Educational Progress (NAEP)

    work page 2022

  3. [3]

    S.D. Sparks. Which nations’ students are defying the math anxiety trend? https://www.edweek.org/ teaching-learning/which-nations-students-are-defying-the-math-anxiety-trend/2024/11 ,

    work page 2024

  4. [4]

    Fernández-Luna, J.F

    J.M. Fernández-Luna, J.F. Huete, A. MacFarlane, and E.N. Efthimiadis. Teaching and learning in information retrieval. Information Retrieval, 12:201–226, 2009

    work page 2009

  5. [5]

    Ott and D

    N. Ott and D. Meurers. Information retrieval for education: Making search engines language aware. Themes in Science and Technology Education, 3:9–30, 2010

    work page 2010

  6. [6]

    Henkel, Z

    O. Henkel, Z. Levonian, C. Li, and M. Postle. Retrieval-augmented generation to improve math question- answering: Trade-offs between groundedness and human preference. In Proceedings of the 17th International Conference on Educational Data Mining , pages 315–320, 2024

    work page 2024

  7. [7]

    Brown, B

    T. Brown, B. Mann, N. Ryder, M. Subbiah, J.D. Kaplan, P. Dhariwal, A. Neelakantan, P. Shyam, G. Sastry, A. Askell, and S. Agarwal. Language models are few-shot learners. Advances in Neural Information Processing Systems, 33:1877–1901, 2020

    work page 1901

  8. [8]

    Chowdhery, S

    A. Chowdhery, S. Narang, J. Devlin, M. Bosma, G. Mishra, A. Roberts, P. Barham, H.W. Chung, C. Sutton, S. Gehrmann, and P. Schuh. Palm: Scaling language modeling with pathways. Journal of Machine Learning Research, 24(240):1–113, 2023

    work page 2023

Show all 51 references
  1. [9]

    Touvron, T

    H. Touvron, T. Lavril, G. Izacard, X. Martinet, M.A. Lachaux, T. Lacroix, B. Rozière, N. Goyal, E. Ham- bro, F. Azhar, and A. Rodriguez. Llama: Open and efficient foundation language models. arXiv preprint arXiv:2302.13971, 2023

    2023 arXiv

  2. [10]

    Jaech, A

    A. Jaech, A. Kalai, A. Lerer, A. Richardson, A. El-Kishky, A. Low, A. Helyar, A. Madry, A. Beutel, A. Carney, and A. Iftimie. Openai o1 system card. arXiv preprint arXiv:2412.16720, 2024

    2024 arXiv

  3. [11]

    D. Guo, D. Yang, H. Zhang, J. Song, R. Zhang, R. Xu, Q. Zhu, S. Ma, P. Wang, X. Bi, and X. Zhang. Deepseek-r1: Incentivizing reasoning capability in llms via reinforcement learning. arXiv preprint arXiv:2501.12948, 2025

    2025 arXiv

  4. [12]

    Achiam, S

    J. Achiam, S. Adler, S. Agarwal, L. Ahmad, I. Akkaya, F.L. Aleman, D. Almeida, J. Altenschmidt, S. Altman, S. Anadkat, and R. Avila. Gpt-4 technical report. arXiv preprint arXiv:2303.08774, 2023

    2023 arXiv

  5. [13]

    Z. Shao, P. Wang, Q. Zhu, R. Xu, J. Song, X. Bi, H. Zhang, M. Zhang, Y .K. Li, Y . Wu, and D. Guo. Deepseekmath: Pushing the limits of mathematical reasoning in open language models. arXiv preprint arXiv:2402.03300, 2024. 13 Machine Assistant with Reliable Knowledge

    2024 arXiv

  6. [14]

    Trinh, Y

    T.H. Trinh, Y . Wu, Q.V . Le, H. He, and T. Luong. Solving olympiad geometry without human demonstrations. Nature, 625(7995):476–482, 2024

    2024

  7. [15]

    Chervonyi, T.H

    Y . Chervonyi, T.H. Trinh, M. Olšák, X. Yang, H. Nguyen, M. Menegali, J. Jung, V . Verma, Q.V . Le, and T. Luong. Gold-medalist performance in solving olympiad geometry with alphageometry2. arXiv preprint arXiv:2502.03544, 2025

    2025

  8. [16]

    Rasul, S

    T. Rasul, S. Nair, D. Kalendra, M. Robin, F. de Oliveira Santini, W.J. Ladeira, M. Sun, I. Day, R.A. Rather, and L. Heathcote. The role of chatgpt in higher education: Benefits, challenges, and future research directions. Journal of Applied Learning and Teaching, 6(1):41–56, 2023

    2023

  9. [17]

    Dempere, K

    J. Dempere, K. Modugu, A. Hesham, and L.K. Ramasamy. The impact of chatgpt on higher education. Frontiers in Education, 8:1206936, September 2023

    2023

  10. [18]

    Mvondo, B

    G.F.N. Mvondo, B. Niu, and S. Eivazinezhad. Exploring the ethical use of llm chatbots in higher education, 2023. Available at SSRN 4548263

    2023

  11. [19]

    S. Dakshit. Faculty perspectives on the potential of rag in computer science higher education. In Proceedings of the 25th Annual Conference on Information Technology Education , pages 19–24, October 2024

    2024

  12. [20]

    Z. Ji, N. Lee, R. Frieske, T. Yu, D. Su, Y . Xu, E. Ishii, Y .J. Bang, A. Madotto, and P. Fung. Survey of hallucination in natural language generation. ACM Computing Surveys, 55(12):1–38, 2023

    2023

  13. [21]

    Kasneci, K

    E. Kasneci, K. Seßler, S. Küchemann, M. Bannert, D. Dementieva, F. Fischer, U. Gasser, G. Groh, S. Günnemann, E. Hüllermeier, and S. Krusche. Chatgpt for good? on opportunities and challenges of large language models for education. Learning and Individual Differences, 103:102274, 2023

    2023

  14. [22]

    Dziri, E

    N. Dziri, E. Kamalloo, S. Milton, O. Zaiane, M. Yu, E.M. Ponti, and S. Reddy. Faithdial: A faithful benchmark for information-seeking dialogue. Transactions of the Association for Computational Linguistics , 10:1473–1490, 2022

    2022

  15. [23]

    L. Zhui, L. Fenghe, W. Xuehu, F. Qining, and R. Wei. Ethical considerations and fundamental principles of large language models in medical education. Journal of Medical Internet Research, 26:e60083, 2024

    2024

  16. [24]

    B. Liu, Z. Cui, S. Hu, X. Li, H. Lin, and Z. Zhang. Llm evaluation based on aerospace manufacturing expertise: Automated generation and multi-model question answering. arXiv preprint arXiv:2501.17183, 2025

    2025 arXiv

  17. [25]

    C. Zhai, S. Wibowo, and L.D. Li. The effects of over-reliance on ai dialogue systems on students’ cognitive abilities: A systematic review. Smart Learning Environments, 11(1):28, 2024

    2024

  18. [26]

    Zhang, S

    P. Zhang, S. Xiao, Z. Liu, Z. Dou, and J.Y . Nie. Retrieve anything to augment large language models. arXiv preprint arXiv:2310.07554, 2023

    2023

  19. [27]

    Shuster, S

    K. Shuster, S. Poff, M. Chen, D. Kiela, and J. Weston. Retrieval augmentation reduces hallucination in conversation. arXiv preprint arXiv:2104.07567, 2021

    2021 arXiv

  20. [28]

    Lewis, E

    P. Lewis, E. Perez, A. Piktus, F. Petroni, V . Karpukhin, N. Goyal, H. Küttler, M. Lewis, W.T. Yih, T. Rocktäschel, and S. Riedel. Retrieval-augmented generation for knowledge-intensive nlp tasks. Advances in Neural Information Processing Systems, 33:9459–9474, 2020

    2020

  21. [29]

    Béchard and O.M

    P. Béchard and O.M. Ayala. Reducing hallucination in structured outputs via retrieval-augmented generation. arXiv preprint arXiv:2404.08189, 2024

    2024 arXiv

  22. [30]

    Hallucination reduction in large language models with retrieval-augmented generation using wikipedia knowledge, 2024

    Jason Kirchenbauer and Caleb Barns. Hallucination reduction in large language models with retrieval-augmented generation using wikipedia knowledge, 2024

    2024

  23. [31]

    C. Niu, Y . Wu, J. Zhu, S. Xu, K. Shum, R. Zhong, J. Song, and T. Zhang. Ragtruth: A hallucination corpus for developing trustworthy retrieval-augmented language models. arXiv preprint arXiv:2401.00396, 2023

    2023 arXiv

  24. [32]

    Y . Gao, Y . Xiong, X. Gao, K. Jia, J. Pan, Y . Bi, Y . Dai, J. Sun, H. Wang, and H. Wang. Retrieval-augmented generation for large language models: A survey. arXiv preprint arXiv:2312.10997, 2:1, 2023

    2023 arXiv

  25. [33]

    D. Thüs, S. Malone, and R. Brünken. Exploring generative ai in higher education: A rag system to enhance student engagement with scientific literature. Frontiers in Psychology, 15:1474892, 2024

    2024

  26. [34]

    Perez, M

    R.M. Perez, M. Shimogawa, Y . Chang, H.A.T. Phan, J.G. Marmorstein, E.S. Yanagawa, and E.J. Petersson. Large language models for education: Chemtask–an open-source paradigm for automated q&a in the graduate classroom. arXiv preprint arXiv:2502.00016, 2025

    2025 arXiv

  27. [35]

    Faghih Shojaei, R

    M. Faghih Shojaei, R. Gulati, B.A. Jasperson, S. Wang, S. Cimolato, D. Cao, W. Neiswanger, and K. Garikipati. Ai-university: An llm-based platform for instructional alignment to scientific classrooms. arXiv e-prints, page arXiv:2504, 2025. 14 Machine Assistant with Reliable Knowledge

    2025

  28. [36]

    I. Ma, A. Krone-Martins, and C. Videira Lopes. Integrating ai tutors in a programming course. In Proceedings of the 2024 on ACM Virtual Global Computing Education Conference V . 1, pages 130–136, December 2024

    2024

  29. [37]

    T. Feng, S. Liu, and D. Ghosal. Courseassist: Pedagogically appropriate ai tutor for computer science education. In Proceedings of the 2024 on ACM Virtual Global Computing Education Conference V . 2, pages 310–311, December 2024

    2024

  30. [38]

    Thway, J

    M. Thway, J. Recatala-Gomez, F.S. Lim, K. Hippalgaonkar, and L.W. Ng. Battling botpoop using genai for higher education: A study of a retrieval augmented generation chatbot’s impact on learning. arXiv preprint arXiv:2406.07796, 2024

    2024 arXiv

  31. [39]

    Schütze, C.D

    H. Schütze, C.D. Manning, and P. Raghavan. Introduction to Information Retrieval , volume 39. Cambridge University Press, Cambridge, 2008

    2008

  32. [40]

    J. Ramos. Using tf-idf to determine word relevance in document queries. In Proceedings of the First Instructional Conference on Machine Learning, volume 242, pages 29–48, December 2003

    2003

  33. [41]

    Robertson and H

    S. Robertson and H. Zaragoza. The probabilistic relevance framework: Bm25 and beyond. F oundations and Trends® in Information Retrieval, 3(4):333–389, 2009

    2009

  34. [42]

    Devlin, M.W

    J. Devlin, M.W. Chang, K. Lee, and K. Toutanova. Bert: Pre-training of deep bidirectional transformers for language understanding. In Proceedings of the 2019 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies,...

    2019

  35. [43]

    Karpukhin, B

    V . Karpukhin, B. Oguz, S. Min, P.S. Lewis, L. Wu, S. Edunov, D. Chen, and W.T. Yih. Dense passage retrieval for open-domain question answering. In EMNLP (1), pages 6769–6781, November 2020

    2020

  36. [44]

    Johnson, M

    J. Johnson, M. Douze, and H. Jégou. Billion-scale similarity search with gpus. IEEE Transactions on Big Data , 7(3):535–547, 2019

    2019

  37. [45]

    Jégou, M

    H. Jégou, M. Douze, and C. Schmid. Product quantization for nearest neighbor search. IEEE Transactions on Pattern Analysis and Machine Intelligence, 33(1):117–128, 2010

    2010

  38. [46]

    L. Gao, Z. Dai, and J. Callan. Coil: Revisit exact lexical match in information retrieval with contextualized inverted list. arXiv preprint arXiv:2104.07186, 2021

    2021 arXiv

  39. [47]

    Sawarkar, A

    K. Sawarkar, A. Mangal, and S.R. Solanki. Blended rag: Improving rag (retriever-augmented generation) accuracy with semantic search and hybrid query-based retrievers. In2024 IEEE 7th International Conference on Multimedia Information Processing and Retrieval (MIPR) , pages 155...

    2024

  40. [48]

    Y . Liu, M. Ott, N. Goyal, J. Du, M. Joshi, D. Chen, O. Levy, M. Lewis, L. Zettlemoyer, and V . Stoyanov. Roberta: A robustly optimized bert pretraining approach. arXiv preprint arXiv:1907.11692, 2019

    1907 arXiv

  41. [49]

    Reimers and I

    N. Reimers and I. Gurevych. Sentence-bert: Sentence embeddings using siamese bert-networks. arXiv preprint arXiv:1908.10084, 2019

    1908 arXiv

  42. [50]

    Jina embeddings 2: 8192-token general-purpose text embeddings for long documents, 2023

    Michael Günther, Jackmin Ong, Isabelle Mohr, Alaeddine Abdessalem, Tanguy Abel, Mohammad Kalim Akram, Susana Guzman, Georgios Mastrapas, Saba Sturua, Bo Wang, Maximilian Werk, Nan Wang, and Han Xiao. Jina embeddings 2: 8192-token general-purpose text embeddings for long docume...

    2023

  43. [51]

    jina-embeddings-v4: Universal embeddings for multimodal multilingual retrieval, 2025

    Michael Günther, Saba Sturua, Mohammad Kalim Akram, Isabelle Mohr, Andrei Ungureanu, Bo Wang, Sedigheh Eslami, Scott Martens, Maximilian Werk, Nan Wang, and Han Xiao. jina-embeddings-v4: Universal embeddings for multimodal multilingual retrieval, 2025. 15

    2025

Pith tools

Reviewed August 6, 2026 · model on record in the stance chip above.