Keywords
LLM
kod generatsiyasi
dasturlash, sun’iy intellekt cheklovlari
gallyusinatsiya, kod xavfsizligi
How to Cite
Abstract
Ushbu maqolada Katta til modellari (LLM) yordamida kod generatsiyasi jarayonining zamonaviy dasturiy ta'minot ishlab chiqishdagi o‘rni, uning asosiy imkoniyatlari va mavjud cheklovlari tahlil qilinadi. Tadqiqotda LLM’larning kod yozish tezligini oshirish, kontekstni tushunish va rutin vazifalarni avtomatlashtirishdagi samaradorligi ko‘rib chiqiladi. Shu bilan birga, generatsiya qilingan kodlardagi xavfsizlik zaifliklari, gallyusinatsiyalar, mualliflik huquqi muammolari va modellarning murakkab arxitekturaviy yechimlardagi cheklovlari chuqur tahlil etiladi. Maqola yakunida LLM’lardan dasturlashda xavfsiz va samarali foydalanish bo‘yicha amaliy tavsiyalar ilgari suriladi
References
1. McKinsey & Company. Unleashing developer productivity with generative AI. 2023.https://www.mckinsey.com/capabilities/mckinsey-digital/ourinsights/unleashing-developer-productivity-with-generative-ai
2. GitHub & Microsoft Research. Measuring AI's Impact on Developer Productivity. arXiv preprint, 2024. https://arxiv.org/abs/2403.18708
3. OpenAI. GPT-4 Technical Report. Journal of Artificial Intelligence Research, 2023. https://cdn.openai.com/papers/gpt-4.pdf
4. GitHub. Does GitHub Copilot Improve Code Quality? GitHub Blog, 2023. https://github.blog/news-insights/research/does-github-copilot-improve-codequality-heres-what-the-data-says
5. Google AI Blog. AlphaCode: Solving Competitive Programming Problems with AI. 2023. https://ai.googleblog.com/2023/01/alphacode.html
This work is licensed under a Creative Commons Attribution 4.0 International License.
