BILIMGA ASOSLANGAN AVTONOM AGENTLAR VA RAG TEXNOLOGIYALARI
DOI:
https://doi.org/10.5281/zenodo.21719601Ключевые слова:
bilim talab qiluvchi agentlar, RAG, katta til modellari, agentli RAG, ko‘p agentli tizimlar.Аннотация
Ushbu maqolada katta til modellarining parametrik xotiraga qaramligi, bilimlarning eskirishi va gallyutsinatsiya
muammolarini kamaytirishga xizmat qiluvchi qidiruv bilan boyitilgan generatsiya (Retrieval-Augmented Generation
— RAG) tizimlari agentli yondashuv asosida tahlil qilinadi. RAG tizimlarining oddiy qidiruv-generatsiya arxitekturasidan
avtonom va ko‘p agentli modellargacha bo‘lgan rivojlanish bosqichlari yoritiladi. Bilim talab qiluvchi agentlar bilimni qo‘lga
kirituvchi va undan foydalanuvchi turlarga ajratilib, ularning so‘rovni qayta ishlash, manbalarni tanlash, dalillarni reytinglash,
umumlashtirish va saqlashdagi vazifalari ko‘rib chiqiladi. Shuningdek, gibrid ko‘p agentli arxitekturalar, agentlar
o‘rtasidagi bilim oqimini muvofiqlashtirish, katta til modellari asosidagi ma’lumotlar sintezi hamda LLM-hakam baholash
mexanizmlari tahlil qilinadi. Agentli RAG katta til modellarining ishonchliligi, moslashuvchanligi va murakkab vazifalarni
bajarish samaradorligini oshiruvchi istiqbolli yo‘nalish sifatida baholanadi.
Библиографические ссылки
Abdul-Jaleel N., Allan J., Croft W.B., Diaz F., Larkey L., Li X., Smucker M.D., Wade C. UMass at TREC 2004: Novelty
and HARD // Proceedings of TREC. – NIST, 2004.
Abdul-Jaleel N., Allan J., Croft W.B., Diaz F., Larkey L., Li X., Smucker M.D., Wade C. UMass at TREC 2004: Novelty
and HARD. – 2004.
Adlakha V., BehnamGhader P., Lu X.H., Meade N., Reddy S. Evaluating Correctness and Faithfulness of Instruction-
Following Models for Question Answering // Transactions of the Association for Computational Linguistics. – 2024. –
Vol. 12. – P. 681–699. DOI: 10.1162/tacl_a_00667.
Aggarwal S., Mandowara D., Agrawal V., Khandelwal D., Singla P., Garg D. Explanations for CommonsenseQA: New
Dataset and Models // Proceedings of the 59th Annual Meeting of the Association for Computational Linguistics and
the 11th International Joint Conference on Natural Language Processing (ACL-IJCNLP 2021). – 2021. – P. 3050–3065.
Alotaibi R., Lei C., Quamar A., Efthymiou V., Özcan F. Property Graph Schema Optimization for Domain-Specific
Knowledge Graphs // Proceedings of the IEEE 37th International Conference on Data Engineering (ICDE 2021). –
– P. 924–935.
Amini A., Gabriel S., Lin S., Koncel-Kedziorski R., Choi Y., Hajishirzi H. MathQA: Towards Interpretable Math Word
Problem Solving with Operation-Based Formalisms // Proceedings of the North American Chapter of the Association
for Computational Linguistics: Human Language Technologies (NAACL-HLT 2019). – 2019. – P. 2357–2367.
An Y., Cheng Y., Park S.J., Jiang J. HyperRAG: Enhancing Quality-Efficiency Tradeoffs in Retrieval-Augmented
Generation with Reranker KV-Cache Reuse // arXiv preprint. – 2025. – arXiv:2504.02921.
Anantha R., Vodianik D. Context Tuning for Retrieval-Augmented Generation // Proceedings of the 1st Workshop on
Uncertainty-Aware NLP (UncertaiNLP 2024). – 2024. – P. 15–22.
Asai A., Wu Z., Wang Y., Sil A., Hajishirzi H. Self-RAG: Learning to Retrieve, Generate, and Critique Through Self-
Reflection // Proceedings of the Twelfth International Conference on Learning Representations (ICLR 2024). – 2024.
Asai A., Zhong Z., Chen D., Koh P.W., Zettlemoyer L., Hajishirzi H., Yih W.-t. Reliable, Adaptable, and Attributable
Language Models with Retrieval // arXiv preprint. – 2024. – arXiv:2403.03187.
Choi N., Lee Y., Cho G.-H., Jeong H., Kong J., Kim S., Park K., Cho S., Jeong I., Nam G., Han S., Yang W., Choi
J. RRADistill: Distilling LLMs’ Passage Ranking Ability for Long-Tail Queries Document Re-Ranking on a Search
Engine // Proceedings of the 2024 Conference on Empirical Methods in Natural Language Processing (EMNLP 2024),
Industry Track. – 2024. – P. 627–641.
Загрузки
Опубликован
Выпуск
Раздел
Лицензия
Copyright (c) 2026 MUHANDISLIK VA IQTISODIYOT
Это произведение доступно по лицензии Creative Commons «Attribution» («Атрибуция») 4.0 Всемирная.