Large Language Models: A Comprehensive Survey of its Applications, Challenges, Limitations, and Future Prospects

Опубликована: Ноя. 16, 2023

<p>Within the vast expanse of computerized language processing, a revolutionary entity known as Large Language Models (LLMs) has emerged, wielding immense power in its capacity to comprehend intricate linguistic patterns and conjure coherent contextually fitting responses. models are type artificial intelligence (AI) that have emerged powerful tools for wide range tasks, including natural processing (NLP), machine translation, question-answering. This survey paper provides comprehensive overview LLMs, their history, architecture, training methods, applications, challenges. The begins by discussing fundamental concepts generative AI architecture pre- trained transformers (GPT). It then an history evolution over time, different methods been used train them. discusses applications medical, education, finance, engineering. also how LLMs shaping future they can be solve real-world problems. challenges associated with deploying scenarios, ethical considerations, model biases, interpretability, computational resource requirements. highlights techniques enhancing robustness controllability addressing bias, fairness, generation quality issues. Finally, concludes highlighting LLM research need addressed order make more reliable useful. is intended provide researchers, practitioners, enthusiasts understanding evolution, By consolidating state-of-the-art knowledge field, this serves valuable further advancements development utilization applications. GitHub repo project available at https://github.com/anas-zafar/LLM-Survey</p>

Язык: Английский

---

Artificial intelligence and consumer behavior: From predictive to generative AI

Journal of Business Research, Год журнала: 2024, Номер 180, С. 114720 - 114720

Опубликована: Май 23, 2024

Язык: Английский

---

Reinforcement Learning for Generative AI: State of the Art, Opportunities and Open Research Challenges

Journal of Artificial Intelligence Research, Год журнала: 2024, Номер 79, С. 417 - 446

Опубликована: Фев. 6, 2024

Generative Artificial Intelligence (AI) is one of the most exciting developments in Computer Science last decade. At same time, Reinforcement Learning (RL) has emerged as a very successful paradigm for variety machine learning tasks. In this survey, we discuss state art, opportunities and open research questions applying RL to generative AI. particular, will three types applications, namely, an alternative way generation without specified objectives; generating outputs while concurrently maximizing objective function; and, finally, embedding desired characteristics, which cannot be easily captured by means function, into process. We conclude survey with in-depth discussion challenges fascinating emerging area.

Язык: Английский

---

The rise and potential of large language model based agents: a survey

Science China Information Sciences, Год журнала: 2025, Номер 68(2)

Опубликована: Янв. 17, 2025

Язык: Английский

---

Synthetic Data Generation with Large Language Models for Text Classification: Potential and Limitations

Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing, Год журнала: 2023, Номер unknown

Опубликована: Янв. 1, 2023

The collection and curation of high-quality training data is crucial for developing text classification models with superior performance, but it often associated significant costs time investment. Researchers have recently explored using large language (LLMs) to generate synthetic datasets as an alternative approach. However, the effectiveness LLM-generated in supporting model inconsistent across different tasks. To better understand factors that moderate data, this study, we look into how performance trained on these may vary subjectivity classification. Our results indicate subjectivity, at both task level instance level, negatively data. We conclude by discussing implications our work potential limitations leveraging LLM generation.

Язык: Английский

---

Navigating the Web of Disinformation and Misinformation: Large Language Models as Double-Edged Swords

IEEE Access, Год журнала: 2024, Номер unknown, С. 1 - 1

Опубликована: Янв. 1, 2024

This paper explores the dual role of Large Language Models (LLMs) in context online misinformation and disinformation. In today's digital landscape, where internet social media facilitate rapid dissemination information, discerning between accurate content falsified information presents a formidable challenge. Misinformation, often arising unintentionally, disinformation, crafted deliberately, are at forefront this LLMs such as OpenAI's GPT-4, equipped with advanced language generation abilities, present double-edged sword scenario. While they hold promise combating by fact-checking detecting LLM-generated text, their ability to generate realistic, contextually relevant text also poses risks for creating propagating misinformation. Further, plagued many problems biases, knowledge cutoffs, hallucinations, which may further perpetuate The outlines historical developments detection how it affects consumption, especially among youth, introduces applications various domains. It then critically analyzes potential counter disinformation sensitive topics healthcare, COVID-19, political agendas. discusses mitigation strategies, ethical considerations, regulatory measures, summarizing previous methods proposing future research direction toward leveraging benefits while minimizing misuse risks. concludes acknowledging powerful tools significant implications both spreading age.

Язык: Английский

---

The Crowdless Future? How Generative AI Is Shaping the Future of Human Crowdsourcing

SSRN Electronic Journal, Год журнала: 2023, Номер unknown

Опубликована: Янв. 1, 2023

This study investigates the capability of generative artificial intelligence (AI) in creating innovative business solutions compared to human crowdsourcing methods. We initiated a challenge focused on sustainable, circular economy opportunities. The attracted diverse range solvers from myriad countries and industries. Simultaneously, we employed GPT-4 generate AI using three different prompt levels, each calibrated simulate distinct crowd expert personas. 145 evaluators assessed randomized selection 10 out 234 solutions, total 1,885 evaluator-solution pairs. Results showed comparable quality between AI-generated solutions. However, ideas were perceived as more novel, whereas delivered better environmental financial value. use natural language processing techniques rich solution text show that although cover similar industries application, exhibit greater semantic diversity. connection diversity novelty is stronger suggesting differences how created by humans or detected evaluators. illuminates potential limitations both solve complex organizational problems sets groundwork for possible integrative human-AI approach problem-solving.

Язык: Английский

---

Exploring Cognitive Strategies in Human-AI Interaction: ChatGPT's Role in Creative Tasks

Journal of Creativity, Год журнала: 2025, Номер unknown, С. 100095 - 100095

Опубликована: Фев. 1, 2025

Язык: Английский

---

The Hard Problems of AI

Опубликована: Фев. 28, 2025

There is currently an enlivened debate regarding the possibility of AI consciousness and/or sentience, as well arguably more partial capabilities we associate with such intelligence or creativity. The itself can be traced back to inception computing, but its current revitalisation powered by recent advancements in field artificial that saw a swift increase act seemingly human-like ways. I argue methodologically flawed, it approaches question consciousness, etc. decidable dealing matters fact. Those engaged are driven desire find suitable definition e.g. would allow them definitively settle whether particular system conscious. However, drawing on Ludwig Wittgenstein’s later philosophy, no exists, because predicates inherently vague (meaning any verdicts they yield bound vague, too). Moreover, impression might directly unobservable fact flawed generalisation practice observation reports sensation reports[1]. In reality, third-person (sentience, agency etc.) attributions independent stipulated internal process happening inside those persons (or systems, case AI). Therefore, only sense which meaningfully asked pragmatic sense: what best _think systems as? _But this subject so sociological and psychological factors, not conceptual ones. Therefore cannot decided aforementioned strategies.

Язык: Английский

---

The Hard Problems of AI

Qeios, Год журнала: 2025, Номер 7(4)

Опубликована: Апрель 10, 2025

There is currently an enlivened debate regarding the possibility of AI consciousness and/or sentience, as well arguably more partial capabilities we associate with such intelligence or creativity. The itself can be traced back to inception computing, but its current revitalisation powered by recent advancements in field artificial that saw a swift increase act seemingly human-like ways. I argue methodologically flawed, it approaches question consciousness, intelligence, etc. decidable dealing matters fact. Those engaged are driven desire find suitable definition e.g. would allow them definitively settle whether particular system conscious. However, drawing on Ludwig Wittgenstein’s later philosophy, no exists, because predicates inherently vague (meaning any verdicts they yield bound vague, too). Moreover, impression might directly unobservable fact flawed generalisation practice observation reports sensation reports[1]. In reality, third-person (sentience, agency, etc.) attributions independent stipulated internal process happening inside those persons (or systems, case AI). Therefore, only sense which meaningfully asked pragmatic sense: what best _think systems as? _But this subject sociological and psychological factors, not conceptual ones. cannot decided aforementioned strategies.

Язык: Английский

---