Pith. sign in

REVIEW 4 major objections 8 minor 38 references

AQUA: A Large Language Model for Aquaculture & Fisheries

T0 review · 4 major / 8 minor · reviewed 2026-08-15 · deepseek-v4-flash

Pith's one-line read AQUA, the first aquaculture-specific LLM, is claimed to outperform general-purpose models on disease, feeding, water, and farm-economics advice.

desk verdict Ships a real aquaculture LLM and a coherent data pipeline, but the central 'consistently outperforms' claim rests on four unblinded, handpicked questions and a circular evaluation. read the letter →

arxiv 2507.20520 v1 pith:WL5OEFES submitted 2025-07-28 cs.CL cs.AIcs.CEcs.LGcs.RO

classification cs.CLcs.AIcs.CEcs.LGcs.RO
keywords AquacultureLargeLanguageModelsArtificialIntelligenceAQUAAQUADAPTSustainabilityAutomationLLM-as-a-Judge
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

AQUA is built by fine-tuning a 7-billion-parameter instruction-tuned model on roughly three million synthetic question–answer pairs produced by AQUADAPT, an agentic pipeline anchored in an expert-built taxonomy of eleven aquaculture categories and over sixty subcategories. The paper's aim is to show that AQUA is the first large language model purpose-built for aquaculture and that this domain-specific training lets it consistently outperform general-purpose instruction models across fish health, feeding and nutrition, production economics, and water quality. The load-bearing design is a fine-tuned LLM judge: calibrated against a gold standard of ten thousand expert-rated pairs, it scores every generated pair and keeps only those rated 4 or above, turning scarce expert judgment into a scalable dataset filter. If the claim holds, expert-level aquaculture advisory can be delivered cheaply and at scale, including on low-power farm devices through a companion small-model IoT system.

What carries the argument

The central machine is AQUADAPT (Data Acquisition, Processing, and Tuning), an agentic data pipeline whose output is the roughly three-million-pair instruction dataset, with the Scoring Agent as the load-bearing component. The Scoring Agent is a fine-tuned GPT-4.1 judge that, prompted with few-shot expert-rated examples, assigns every QA pair a score from 2 to 5 and filters out anything below 4; the paper argues this transfer is sound because the judge outperforms two alternatives on agreement (Spearman $\rho = 0.85$, Kendall $\tau = 0.79$, Pearson $r = 0.89$; pairwise consistency 88.5%; weighted Cohen's $\kappa = 0.76$). Domain specialization is then applied with low-rank adapters on the surviving pairs.

What would settle it

Have independent aquaculture experts blindly rate a random sample of pairs the judge rejected (scores below 4) and a matching sample it accepted, then compare rating distributions: substantial numbers of 4-or-above scores among rejected pairs, or below-4 scores among accepted pairs, would falsify the judge's generalization. The complementary test is a fresh expert-written question set, independent of AQUADAPT-generated data, on which AQUA and the general-purpose baselines are rated blind.

Watch

Extended reading notes

Core claim

On the paper's own terms, the discovery is that expert-anchored, judge-filtered synthetic data is enough to make a compact 7-billion-parameter model answer aquaculture questions at a level a rubric-based comparison treats as expert. AQUA receives an average rating of 4.75 out of 5 across four topical areas, versus 3.25 for Llama-3.1-8B and 2.75 for Qwen-2.5-7B, and its answers reproduce expert norms such as feed conversion ratios near 1.5, daily monitoring intervals, and dissolved-oxygen thresholds above 5 mg/L. The claimed novelty is the AQUADAPT agentic framework: a Data Agent cleans 55,105 documents; an Expert Agent supplies seed question–answer pairs and the category taxonomy; a QA Agent generates pairs from prompts and from BM25-filtered literature; and a Scoring Agent applies a fine-tuned GPT-4.1 judge, chosen because it best matches expert ratings (Spearman $\rho = 0.85$, Kendall $\tau = 0.79$, Pearson $r = 0.89$, MAE 0.42), to admit only pairs scoring 4 or above.

Load-bearing premise

The load-bearing premise is that the judge model scores the roughly three million question–answer pairs as faithfully as experts scored the ten thousand gold-standard pairs, so the filtered training data — and anything learned from it — is genuinely expert-level.

Editorial extensions

If this is right

  • AQUA could serve as a low-cost advisory layer for fish farmers and hatchery operators, supplying procedural answers with concrete intervals, thresholds, and corrective actions.
  • The AQUADAPT recipe — expert seeds, dual-path synthetic QA generation, and a judge-model filter calibrated to a small gold standard — is portable to other specialized domains where expert annotation is scarce.
  • A 1-billion-parameter version of AQUA can run on edge hardware such as a Raspberry Pi, using MQTT sensor feeds to generate real-time water-quality alerts and corrective recommendations without internet access.
  • Domain specialization measurably improves factual precision over general models in the paper's comparisons, such as correct feed-conversion ratios and global seafood-production shares where generalist baselines give wrong values.
  • The expert-calibrated LLM judge establishes a template for scalable quality control of synthetic instruction data in future domain-specific fine-tuning.

Reading between the lines

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

  • A direct ablation the paper does not report — training on the unfiltered pool versus the judge-filtered subset — would test whether threshold filtering itself is what creates the expert-level answers, not just the volume of domain text.
  • Because the held-out validation set and the four illustrative comparisons come from the same generation pipeline, the sharper test is an independent expert-written benchmark; until that exists, part of AQUA's apparent edge may be label agreement with its own filter.
  • The IoT scenario suggests that once on-device advisory is reliable, the practical bottleneck in real farms shifts to sensor calibration, data latency, and alarm fatigue rather than model quality.
  • If the judge-generalization assumption holds across domains, the AQUADAPT pipeline implies that small, cheap models can rival much larger general models in any vertical with a willing expert community and enough open-access literature.
Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

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

Referee Report

4 major / 8 minor

Summary. The paper presents AQUA, a 7B-parameter instruction-tuned model for aquaculture, built by fine-tuning Mistral-7B-Instruct-v0.3 with LoRA on approximately 3 million question-answer pairs generated by the AQUADAPT agentic pipeline. The pipeline combines web-scraped literature, expert-curated seed questions, dual-path generation using fine-tuned GPT-4.1 and Gemini 2.0 Flash, BM25 filtering, and an LLM-as-a-judge filtering stage calibrated on 10,000 expert-rated examples. The authors claim that AQUA is the first aquaculture-specific LLM and that it "consistently outperformed" general-purpose models such as Qwen-2.5-7B and Llama-3.1-8B. The evaluation consists of BLEU/ROUGE scores on a 20,000-pair held-out set and a four-question expert-rated comparison spanning fish health, feeding, economics, and water quality.

Significance. If the performance claims were supported, AQUA would be a useful community resource: the authors release model checkpoints and test data, document training hyperparameters, and describe a modular data-generation framework that could be adapted to other specialized domains. The expert-in-the-loop seed curation and judge-calibration strategy is a reasonable methodological direction. However, the current evidence does not support the central claims. The headline comparison rests on four handpicked questions, the held-out evaluation is drawn from the same synthetic pipeline that produced the training data, and no comparison to the base Mistral model is reported. The claimed "expert-level reasoning" is therefore unsubstantiated, even though the limitations section shows appropriate awareness of domain-anchoring risks.

major comments (4)
  1. [Section VIII; Fig. 4] The claim in Section IX that AQUA "consistently outperformed general-purpose instruction models" is supported only by the four hand-constructed questions in Section VIII. No sampling frame, repeated trials, confidence intervals, significance test, or information about rater blinding is reported, and the average ratings (AQUA 4.75, Llama 3.25, Qwen 2.75) are computed over exactly four responses. This evidence cannot establish a distributional superiority claim.
  2. [Section VII-B; Table 4] The BLEU-4 and ROUGE evaluation is conducted on a held-out set of QA pairs generated by the same AQUADAPT pipeline used to create the training data. Without comparison to the base Mistral-7B-Instruct-v0.3 checkpoint or to general-purpose models, these scores primarily measure fidelity to the synthetic reference distribution and do not demonstrate expert-level reasoning. The paper should include baseline comparisons and, ideally, an independently authored test set.
  3. [Section VI] The evaluation loop is circular: the fine-tuned GPT-4.1 judge is selected for agreement with 10,000 expert ratings (Section VI-C), then applied to filter the entire 3-million-pair dataset (Section VI-D); AQUA is fine-tuned on the filtered set (Section VII-A) and evaluated on a held-out set drawn from the same pipeline (Section VII-B). This design bakes the judge's preferences into both training and evaluation. An external validation set with expert labels, generated independently of the AQUADAPT filter, is needed to break the loop.
  4. [Section VI-D] The paper reports that approximately 3 million QA pairs passed the judge's threshold of 4 but does not report the retention rate, the distribution of judge scores, or any estimate of judge reliability on the full corpus, as opposed to the 10,000-pair gold standard. Without this information, the assertion that the judge's agreement with experts generalizes to the whole dataset is unsupported.
minor comments (8)
  1. [Section III] The tool name "Dockling" appears to be a typo for "Docling" (reference [21]).
  2. [Section V-B] The BM25 threshold tau is described as tuned, but its value is not reported; please provide the value or describe the tuning procedure.
  3. [Section II] The related-work paragraph cites "Li et al. (2024)" for OceanGPT, but reference [19] is "Deng et al. (2023)"; please correct the citation.
  4. [Section X] The sentence "see extracted reference on AQUA-1B with Raspberry Pi and MQTT" does not point to any entry in the reference list; either add the reference or remove the pointer.
  5. [References] Reference [15] is listed with the same page range as reference [11] and appears to describe a different work; please verify the bibliographic details.
  6. [Table 1] The two source quantities sum to 55,105, but the collection strategy suggests possible overlap between "Web" and "Open-access Source"; please clarify whether the categories are disjoint.
  7. [Section III; Fig. 1] The caption of Fig. 1 says "Aquaculture Categories" but the figure is not visible in the submitted text; please ensure the figure is included.
  8. [Abstract and Introduction] The novelty claim "first large language model tailored for aquaculture" is stated without a systematic comparison to prior domain-adapted agricultural or marine models; consider softening the claim.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the AQUA pipeline is self-contained and each stage is validated against external expert inputs rather than reducing to its own outputs.

full rationale

The paper's claimed derivation of AQUA from AQUADAPT is not circular by construction. Expert seed QA pairs and the 10,000-pair gold standard are external human inputs; the fine-tuned GPT-4.1 judge is selected by measured agreement with those expert ratings and then applied to filter the synthetic corpus; AQUA is fine-tuned on the filtered data and evaluated on held-out references and expert-rated comparisons. The judge's agreement metrics (rho = 0.85, kappa = 0.76, etc.) are computed against expert labels, not against the judge's own scores, so the selection step has independent content. The held-out BLEU/ROUGE evaluation uses references from the same AQUADAPT pipeline, which is a validity limitation rather than a definitional equivalence. The headline outperformance claim rests on four handpicked questions with no significance testing, which is an evidentiary weakness, not a circular reduction: the comparison is made against general-purpose models using a separate expert rubric, not derived from AQUA's training objective. No load-bearing self-citation appears; the Hugging Face links are pointers to released artifacts, and the passing mention of AQUA-1B in Section X is illustrative, not used to justify the central results. Overall, the derivation chain does not equate any predicted quantity with its fitting input.

Assumptions & free parameters 1 free parameters · 5 assumptions · 2 invented entities

The central claim rests on a pipeline that assumes expert ratings are reliable, generative LLMs produce correct domain content, and the LLM judge generalizes beyond its training gold standard. The BM25 threshold tau is a free parameter not disclosed. The only entity with independent evidence is the model itself, but the paper does not provide an external benchmark.

free parameters (1)
  • BM25 relevance threshold tau = not reported
    Section V-B describes a 'tuned threshold tau' that decides which documents enter the QA generation stage; its value is never given, so the data selection is not reproducible.
assumptions (5)
  • domain assumption Expert ratings on the gold standard are accurate and reliable for quality evaluation.
    The entire pipeline depends on human experts' 2-to-5 star ratings being a correct measure of QA quality; Section IV-B and VI-A. There is no evidence of inter-rater reliability.
  • domain assumption GPT-4.1 and Gemini 2.0 Flash generate factually correct and domain-relevant QAs from few-shot prompts.
    The synthetic data pipeline assumes the generative LLMs produce high-quality content that matches expert knowledge; Section IV-B and V-C.
  • domain assumption The fine-tuned GPT-4.1 judge's scores on the full dataset accurately emulate expert ratings beyond the 10k gold standard.
    Section VI-D applies the judge to all QA pairs with a threshold of 4; if the judge is miscalibrated on out-of-distribution pairs, the final dataset quality is unknown.
  • domain assumption BM25 with the given formula and tuned threshold selects documents relevant to aquaculture.
    Section V-B uses BM25 to filter the corpus; the selection is only as good as the query terms and the threshold, which is not disclosed.
  • standard math BLEU and ROUGE are valid measures of the quality of generated aquaculture responses.
    Section VII-B uses these metrics as evaluation; they are standard but are known to correlate poorly with semantic correctness, especially for open-ended advisory answers.
invented entities (2)
  • AQUA (AQUA-7B) independent evidence
    purpose: Domain-specialized LLM for aquaculture advisory and decision support.
    The model weights are hosted on HuggingFace, so third parties can download and run inference, but the paper provides no independent benchmark to verify the claimed superiority.
  • AQUADAPT framework
    purpose: Agentic pipeline for generating and filtering synthetic instruction data.
    AQUADAPT is a methodology described in prose; no code is released, and its components (Data Agent, Expert Agent, etc.) cannot be independently tested without implementation.

how reviews work

0 comments
Cite this review

Pith. "Pith review of AQUA: A Large Language Model for Aquaculture & Fisheries." pith.science (2026). https://pith.science/paper/WL5OEFES

@misc{pith2026250720520,
  author       = {Pith},
  title        = {Pith review of: AQUA: A Large Language Model for Aquaculture & Fisheries},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/WL5OEFES}},
  note         = {Machine review of arXiv:2507.20520}
}
read the original abstract

Aquaculture plays a vital role in global food security and coastal economies by providing sustainable protein sources. As the industry expands to meet rising demand, it faces growing challenges such as disease outbreaks, inefficient feeding practices, rising labor costs, logistical inefficiencies, and critical hatchery issues, including high mortality rates and poor water quality control. Although artificial intelligence has made significant progress, existing machine learning methods fall short of addressing the domain-specific complexities of aquaculture. To bridge this gap, we introduce AQUA, the first large language model (LLM) tailored for aquaculture, designed to support farmers, researchers, and industry practitioners. Central to this effort is AQUADAPT (Data Acquisition, Processing and Tuning), an Agentic Framework for generating and refining high-quality synthetic data using a combination of expert knowledge, largescale language models, and automated evaluation techniques. Our work lays the foundation for LLM-driven innovations in aquaculture research, advisory systems, and decision-making tools.

Figures

Figures reproduced from arXiv: 2507.20520 by the authors.

Figure 1
Figure 1. Aquaculture Categories • Nutrition, Feeding, and Feed Technology: Fo￾cuses on what fish need to eat, how to make good feed, balancing protein and energy, feed ef￾ficiency (FCR), using probiotics and additives, al￾ternative proteins, live food sources, and modern feed-making methods. Good feeding practices can improve growth and sustainability. • Water Quality and Environmental Control: In￾cludes managing water tempe… view at source ↗
Figure 2
Figure 2. Aquaculture response quality ranges from basic to comprehensive. C. Formal Algorithmic Definition Let: • C = {c1, c2, . . . , cn}: Set of aquaculture categories • Pci : Prompt template for category ci • Qseed ci : Expert-authored QA seeds for ci • G: Generative language model (e.g., GPT-4.1) • T: Minimum acceptable score threshold, T = 4 The Expert Agent operates as follows: Algorithm: Expert Agent Instruction Curat… view at source ↗
Figure 3
Figure 3. AQUADAPT Framework • Dfinal QA denotes the final curated set of high-quality question–answer pairs used to train AQUA. • qi represents an individual QA pair from the com￾bined literature and synthetic dataset. • S (L) i is the quality rating assigned by the LLM￾as-a-Judge (implemented within Scoring Agent), based on few-shot prompting calibrated with expert￾rated examples. This process resulted in a final high-integ… view at source ↗
Figures from the paper (1 more)
Figure 4
Figure 4. Figure 4: AQUA consistently achieves higher ratings, with an average of 4.75, followed by Llama at 3.25 and Qwen at 2.75. Try AQUA-7B at Huggingface: https://huggingface.co/KurmaAI/AQUA-7B Each topic was represented by a carefully constructed question grounded in realistic opera…

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

38 extracted references · 30 canonical work pages

  1. [1]

    NOAA Fisheries. (2023). Aquaculture in the United States. U.S. Department of Commerce. [Online]. Available: https://www.fisheries.noaa.gov

  2. [2]

    USDA Economic Research Service. (2022). Aquaculture Data. U.S. Department of Agriculture. [Online]. Available: https://www.ers.usda.gov/data-products/aquaculture-data/

  3. [3]

    Knapp, G., & Rubino, M. C. (2016). The political economics of marine aquaculture in the United States. Reviews in Fisheries Science & Aquaculture , 24(3), 213–229

  4. [4]

    L., & Stone, N

    Engle, N. L., & Stone, N. (2013). Competitiveness of U.S. aquaculture within the current U.S. regulatory framework. Aqua- culture Economics & Management , 17(3), 251–280

  5. [5]

    L., Hanson, C., & Hinshaw, D

    Engle, N. L., Hanson, C., & Hinshaw, D. S. (2021). Cost drivers and profitability of U.S. aquaculture farms. North American Journal of Aquaculture , 83(3), 251–262

  6. [6]

    R., & Tyler, A

    Knapp, G., Anderson, C. R., & Tyler, A. M. (2022). Aquaculture in the United States: Constraints, opportunities, and outlook. Marine Policy, 135, Article 104891

  7. [7]

    Mardones, J., Perez, D., & Rojas, J. (2018). The impact of disease outbreaks on the economics of aquaculture: A case study of infectious salmon anemia in Chile. Aquaculture Economics & Management, 22(2), 152–168

  8. [8]

    A., Hansen, A

    Torrissen, O., Olsen, P. A., Hansen, A. G., et al. (2011). Sustain- able aquaculture: Feeding and environmental impact. Aquaculture Research, 42(S1), 89–101

Show all 38 references
  1. [9]

    L., Peng, H., & Barton, M

    Akinbowale, O. L., Peng, H., & Barton, M. D. (2007). Antimi- crobial resistance in bacteria isolated from aquaculture sources in Australia. Journal of Applied Microbiology , 102(3), 879–885

  2. [10]

    H ˚astein, T., Lillehaug, T., & Jarp, T. (2005). Risk factors as- sociated with infectious salmon anaemia (ISA) outbreaks in Norwegian salmon farming. Diseases of Aquatic Organisms , 64(1), 11–20

  3. [11]

    Moor, M., Banerjee, A., Abad, C., et al. (2023). Foundation models for generalist medical artificial intelligence. Nature, 616, 259–265

  4. [12]

    Deng, H., Chen, B., Chen, M., et al. (2023). Artificial intelligence in Earth science: A review of current advances and future challenges. Earth-Science Reviews, 239, Article 104321

  5. [13]

    Bi, J., Liu, X., Chen, Y ., & Liu, H. (2023). AI-based smart aqua- culture: Challenges, advances, and future directions. Computers and Electronics in Agriculture , 209, Article 107517

  6. [14]

    P., Wani, S

    Kumar, R., Singh, V . P., Wani, S. A., & Singh, B. K. (2022). A comprehensive survey on artificial intelligence and IoT-based systems for smart aquaculture. Aquacultural Engineering , 96, Article 102257

  7. [15]

    Singhal, A., et al. (2023). Large language models encode clinical knowledge. Nature, 616, 259–265

  8. [16]

    Patel, J., Awasthi, M., & Sharma, A. (2023). AgriGPT: Domain adaptation of generative pre-trained transformers for agricultural decision support. Computers and Electronics in Agriculture, 205, Article 107623

  9. [17]

    Luo, R., et al. (2022). BioGPT: Generative pre-trained trans- former for biomedical text generation and mining. Briefings in Bioinformatics, 23(6), 1–11

  10. [18]

    [Online]

    AI4Finance-Foundation, ”FinGPT: An Open-Source Financial Large Language Model Platform,” GitHub, 2024. [Online]. Avail- able: https://github.com/AI4Finance-Foundation/FinGPT

  11. [19]

    Deng, H., et al. (2023). OceanGPT: Foundation models for ocean science. arXiv preprint, arXiv:2308.13262

  12. [20]

    W. Zhou, X. Liu, Y . Chen, and Z. Wang, ”IPM-AgriGPT: A large language model for pest and disease management with a G-EA framework and agricultural contextual reasoning,” Math- ematics, vol. 13, no. 4, p. 566, 2025. [Online]. Available: https://doi.org/10.3390/math13040566

  13. [21]

    B., Omenetti, M., Lindlbauer, F., Dinkla, K., Mishra, L., Kim, Y ., Gupta, S., de Lima, R

    Auer, C., Lysak, M., Nassar, A., Dolfi, M., Livathinos, N., Vagenas, P., Ramis, C. B., Omenetti, M., Lindlbauer, F., Dinkla, K., Mishra, L., Kim, Y ., Gupta, S., de Lima, R. T., Weber, V ., Morin, L., Meijer, I., Kuropiatnyk, V ., & Staar, P. W. J. (2024, December 9). Docling ...

  14. [22]

    S., & Agarwal, S

    Adhikari, N. S., & Agarwal, S. (2025, April 3). A comparative study of PDF parsing tools across diverse document categories. arXiv.org. https://arxiv.org/abs/2410.09871

  15. [23]

    P., Narsale, S

    Mohale, H. P., Narsale, S. A., Baraiya, R., Kadam, R. V ., Prakash, P., Sheikh, S., Mansukhbhai, C. R., & Kir- tikumar, P. B. (2024). Artificial Intelligence in Fisheries and Aquaculture: Enhancing Sustainability and Productivity. Archives of Current Research International, 24...

  16. [24]

    W., Afsarimanesh, N., Alahi, M

    Tina, F. W., Afsarimanesh, N., Alahi, M. E. E., & Nag, A. (2025). Integrating AIoT Technologies in Aqua- culture: A Systematic Review. Future Internet, 17(5), 199. https://doi.org/10.3390/fi17050199

  17. [25]

    S., Kumar, P

    Dennison, M. S., Kumar, P. S., Wamyil, F., Meji, M. A., & Ganapathy, T. (2025). The role of automation and robotics in transforming hydroponics and aquaponics to large scale. Discover Sustainability, 6(1). https://doi.org/10.1007/s43621-025-00908-4

  18. [26]

    H., Van Doan, H., Ashour, M., Davies, S., & El-Haroun, E

    Ragab, S., Rossi, W., Hoseinifar, S. H., Van Doan, H., Ashour, M., Davies, S., & El-Haroun, E. (2025). Overview of Aquaculture Artificial Intelligence (AAI) Applications: Enhance Sustainability and Productivity, Reduce Labor Costs, and Increase the Quality of Aquatic Products....

  19. [27]

    Huang, Y .-P., & Khabusi, S. P. (2025). Artificial Intelligence of Things (AIoT) Advances in Aquaculture: A Review. Processes, 13(1), 73. https://doi.org/10.3390/pr13010073

  20. [28]

    Baena-Navarro, R., Carriazo-Regino, Y ., Pinedo-L ´opez, J., & Torres-Hoyos, F. (2025). Intelligent Prediction and Continuous Monitoring of Water Quality in Aquaculture: Integration of Machine Learning and Internet of Things for Sustainable Man- agement. Water, 17(1), 82. http...

  21. [29]

    I., Mohammed, H., & Ahammad, F

    Islam, S. I., Mohammed, H., & Ahammad, F. (2024). Cutting- edge technologies for detecting and controlling fish diseases: Current status, outlook, and challenges. Journal of the World Aquaculture Society, 55(2). https://doi.org/10.1111/jwas.13051

  22. [30]

    A., V on Davier, M., Lottridge, S., & Harris, D

    Hao, J., Yaneva, V ., V on Davier, A. A., V on Davier, M., Lottridge, S., & Harris, D. J. (2024). Transforming Assessment: The Im- pacts and Implications of Large Language Models and Generative AI. Educational Measurement: Issues and Practice, 43(2), 16–29. https://doi.org/10....

  23. [31]

    Yang, X., Zhang, S., Dong, S., Gao, Q., Liu, J., & Zhou, C. (2020). Deep learning for smart fish farming: applications, opportunities and challenges. Reviews in Aquaculture, 13(1), 66–90. https://doi.org/10.1111/raq.12464

  24. [32]

    A., Huang, C.-T., & Cheng, S.-C

    Lan, H.-Y ., Lin, S.-S., Ubina, N. A., Huang, C.-T., & Cheng, S.-C. (2022). Digital Twin Architecture Evalua- tion for Intelligent Fish Farm Management Using Modified Analytic Hierarchy Process. Applied Sciences, 13(1), 141. https://doi.org/10.3390/app13010141

  25. [33]

    A., Chen, H.-Y ., Cheng, S.-C., Chang, C.-C., & Lan, H.-Y

    Ubina, N. A., Chen, H.-Y ., Cheng, S.-C., Chang, C.-C., & Lan, H.-Y . (2021). A Visual Aquaculture System Us- ing a Cloud-Based Autonomous Drones. Drones, 5(4), 109. https://doi.org/10.3390/drones5040109

  26. [34]

    J., Shen, Y ., Wallis, P., Allen-Zhu, Z., Li, Y ., Wang, S., Wang, L., & Chen, W

    Hu, E. J., Shen, Y ., Wallis, P., Allen-Zhu, Z., Li, Y ., Wang, S., Wang, L., & Chen, W. (2021, October 16). Lora: Low-rank adaptation of large language models. arXiv.org. https://arxiv.org/abs/2106.09685

  27. [35]

    Q., Sablayrolles, A., Mensch, A., Bamford, C., Chaplot, D

    Jiang, A. Q., Sablayrolles, A., Mensch, A., Bamford, C., Chaplot, D. S., Casas, D. de las, Bressand, F., Lengyel, G., Lample, G., Saulnier, L., Lavaud, L. R., Lachaux, M.-A., Stock, P., Scao, T. L., Lavril, T., Wang, T., Lacroix, T., & Sayed, W. E. (2023, Oc- tober 10). Mistra...

  28. [36]

    Qwen, Yang, A., Yang, B., Zhang, B., Hui, B., Zheng, B., Yu, B., Li, C., Liu, D., Huang, F., Wei, H., Lin, H., Yang, J., Tu, J., Zhang, J., Yang, J., Yang, J., Zhou, J., Lin, J., . . . Qiu, Z. (2025, January 3). QWEN2.5 technical report. arXiv.org. https://arxiv.org/abs/2412.15115

  29. [37]

    Papineni, K., Roukos, S., Ward, T., & Zhu, W.-J. (n.d.). Bleu: A method for automatic evaluation of Machine Translation. ACL Anthology. https://aclanthology.org/P02-1040/

  30. [38]

    N., Kaiser, L., & Polosukhin, I

    Vaswani, A., Shazeer, N., Parmar, N., Uszkoreit, J., Jones, L., Gomez, A. N., Kaiser, L., & Polosukhin, I. (2023, August 2). Attention is all you need. arXiv.org. https://arxiv.org/abs/1706.03762

Pith tools

Reviewed August 15, 2026 · model on record in the stance chip above.