Large Language Model-Based Autonomous Agents |
||
 |
 |
|
| © 2024 by IJCTT Journal | ||
| Volume-72 Issue-5 |
||
| Year of Publication : 2024 | ||
| Authors : Prerak Garg, Divya Beeram | ||
| DOI : 10.14445/22312803/IJCTT-V72I5P118 | ||
How to Cite?
Prerak Garg, Divya Beeram, "Large Language Model-Based Autonomous Agents ," International Journal of Computer Trends and Technology, vol. 72, no. 5, pp. 151-162, 2024. Crossref, https://doi.org/10.14445/22312803/IJCTT-V72I5P118
Abstract
Artificial Intelligence (AI) agents represent a significant paradigm shift, offering novel methodologies to enhance efficiency and productivity across functions and industries. This study introduces a novel architectural framework for AI agents, with a focus on leveraging Large Language Models (LLMs), such as OpenAI’s GPT-4, to create a foundation for advanced autonomous functionality. The architecture is designed to enhance the adaptability, efficiency, and intelligence of AI agents across various domains, with software development serving as a primary case study. The framework delineates critical components including the integration mechanism for LLMs, task execution protocols, continuous learning processes, and interaction models, aiming to facilitate the seamless incorporation of AI agents into complex environments. By applying this framework within a software development context, the research demonstrates significant improvements: a 50% reduction in debugging time, a 75% decrease in version control conflicts, and a 35% increase in coding standards compliance. These outcomes not only validate the framework’s effectiveness in real-world applications but also underscore its potential to revolutionize the capabilities of AI agents beyond software development. The proposed architecture promises to empower AI agents with a higher degree of autonomy and intelligence, making them invaluable assets in tackling diverse challenges. The implications of this work are vast, setting a new benchmark for AI agent design and deployment and opening avenues for future research and development in AI-driven innovations.
Keywords
AI Agents, AI-enabled Software Development, Artificial Intelligence, Autonomous Agents, LLM Agents.
Reference
[1] Tom B. Brown et al., “Language Models are Few-Shot Learners,” arXiv, pp. 1-75, 2020.
[CrossRef] [Google Scholar] [Publisher Link]
[2] Jacob Devlin et al., “BERT: Pre-Training of Deep Bidirectional Transformers for Language Understanding, arXiv, pp. 4171-4186, 2019.
[CrossRef] [Google Scholar] [Publisher Link]
[3] Cari Beth Head et al., “Large Language Model Applications for Evaluation: Opportunities and Ethical Implications,” New Directions for Evaluation, vol. 2023, no. 178-179, pp. 33-46, 2023.
[CrossRef] [Google Scholar] [Publisher Link]
[4] Chenxu Hu et al., “Chatdb: Augmenting LLMs with Databases as Their Symbolic Memory,” arXiv, pp. 1-12, 2023.
[CrossRef] [Google Scholar] [Publisher Link]
[5] Guanzhi Wang et al., “Voyager: An Open-Ended Embodied Agent with Large Language Models,” arXiv, pp. 1-42, 2023.
[CrossRef] [Google Scholar] [Publisher Link]
[6] Suchin Gururangan et al., “Don’t Stop Pretraining: Adapt Language Models to Domains and Tasks,” arXiv, pp. 1-19, 2020.
[CrossRef] [Google Scholar] [Publisher Link]
[7] Hongru Wang et al., “Chain-of-Thought Prompting for Responding to In-Depth Dialogue Questions with LLMs,” arXiv, pp. 1-18, 2023.
[CrossRef] [Google Scholar] [Publisher Link]
[8] Alon Halevy, Peter Norvig, and Fernando Pereira, “The Unreasonable Effectiveness of Data,” IEEE Intelligent Systems, vol. 24, no. 2, pp. 8-12, 2009.
[CrossRef] [Google Scholar] [Publisher Link]
[9] Jeremy Howard, and Sebastian Ruder, “Universal Language Model Fine-Tuning for Text Classification,” arXiv, pp. 1-12, 2018.
[CrossRef] [Google Scholar] [Publisher Link]
[10] Javier Insa-Cabrera et al., “Comparing Humans and AI Agents,” Artificial General Intelligence, Lecture Notes in Computer Science, Mountain View, CA, USA, vol. 6830, pp. 122-132, 2011.
[CrossRef] [Google Scholar] [Publisher Link]
[11] Kawin Ethayarajh, “How Contextual are Contextualized Word Representations? Comparing the Geometry of BERT, ELMo, and GPT2 Embeddings,” arXiv, pp. 1-11, 2019.
[CrossRef] [Google Scholar] [Publisher Link]
[12] Kostas Hatalis et al., “Memory Matters: The Need to Improve Long-Term Memory in LLM-Agents,” Proceedings of the AAAI Fall Symposium Series, vol. 2, no. 1, pp. 277-280, 2023.
[CrossRef] [Google Scholar] [Publisher Link]
[13] Jared Kaplan et al., “Scaling Laws for Neural Language Models,” arXiv, pp. 1-30, 2020.
[CrossRef] [Google Scholar] [Publisher Link]
[14] Sergey Levine et al., “End-to-End Training of Deep Visuomotor Policies,” Journal of Machine Learning Research, vol. 17, no. 39, pp. 1-40, 2016.
[CrossRef] [Google Scholar] [Publisher Link]
[15] Liunian Harold Li et al., “Visualbert: A Simple and Performant Baseline for Vision and Language,” arXiv, pp. 1-14, 2019.
[CrossRef] [Google Scholar] [Publisher Link]
[16] Jelena Luketina et al., “A Survey of Reinforcement Learning Informed by Natural Language,” arXiv, pp. 1-9, 2019.
[CrossRef] [Google Scholar] [Publisher Link]
[17] Victor Sanh et al., “DistilBERT, A Distilled Version of BERT: Smaller, Faster, Cheaper and Lighter,” arXiv, pp. 1-5, 2019.
[CrossRef] [Google Scholar] [Publisher Link]
[18] Reza Shokri, and Vitaly Shmatikov, “Privacy-Preserving Deep Learning,” Proceedings of the 22nd ACM SIGSAC Conference on Computer and Communications Security, pp. 1310-1321, 2015.
[CrossRef] [Google Scholar] [Publisher Link]
[19] Mohit Shridhar et al., “ALFRED: A Benchmark for Interpreting Grounded Instructions for Everyday Tasks,” 2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), Seattle, WA, USA, pp. 10740-10749, 2020.
[CrossRef] [Google Scholar] [Publisher Link]
[20] Tiziano Labruna et al., “Unraveling Chatgpt: A Critical Analysis of AI-Generated Goal-Oriented Dialogues and Annotations,” International Conference of the Italian Association for Artificial Intelligence, Rome, Italy, pp. 151-171, 2023.
[CrossRef] [Google Scholar] [Publisher Link]
[21] Varun Nair et al., “DERA: Enhancing Large Language Model Completions with Dialog-Enabled Resolving Agents,” arXiv, pp. 1-38, 2023.
[CrossRef] [Google Scholar] [Publisher Link]
[22] Ashish Vaswani et al., “Attention is all you need,” Advances in Neural Information Processing Systems 30, pp. 1-15, 2017.
[CrossRef] [Google Scholar] [Publisher Link]
[23] Wanjun Zhong et al., “Memorybank: Enhancing Large Language Models with Long-Term Memory,” Proceedings of the AAAI Conference on Artificial Intelligence, vol. 38, no. 17, pp. 19724-19731, 2024.
[CrossRef] [Google Scholar] [Publisher Link]
[24] Zhiheng Xi et al., “The Rise and Potential of Large Language Model Based Agents: A Survey,” arXiv, pp. 1-86, 2023.
[CrossRef] [Google Scholar] [Publisher Link]
[25] Samuel Benton et al., “Evaluating and Improving Unified Debugging,” IEEE Transactions on Software Engineering, vol. 48, no. 11, pp. 4692-4716, 2021.
[CrossRef] [Google Scholar] [Publisher Link]
[26] Pilar Rodríguez et al., “Advances in using Agile and Lean Processes for Software Development,” Advances in Computers, vol. 113, pp. 135-224, 2019.
[CrossRef] [Google Scholar] [Publisher Link]
[27] James V. Wertsch, and Richard Sohmer, “Vygotsky on Learning and Development,” Human Development, vol. 38, no. 6, pp. 332-337, 1995.
[CrossRef] [Google Scholar] [Publisher Link]
[28] Tushar Chugh et al., “Intelligent Agents Driven Data Analytics using Large Language Models,” 2023 International Conference on Artificial Intelligence, Blockchain, Cloud Computing, and Data Analytics (ICoABCD), Denpasar, Indonesia, pp. 152-157, 2023.
[CrossRef] [Google Scholar] [Publisher Link]