Enhancing Genetic Improvement Mutations Using Large Language Models DOI Creative Commons
Alexander E. I. Brownlee, James P. Callan, Karine Even-Mendoza

et al.

arXiv (Cornell University), Journal Year: 2023, Volume and Issue: unknown

Published: Jan. 1, 2023

Large language models (LLMs) have been successfully applied to software engineering tasks, including program repair. However, their application in search-based techniques such as Genetic Improvement (GI) is still largely unexplored. In this paper, we evaluate the use of LLMs mutation operators for GI improve search process. We expand Gin Java toolkit call OpenAI's API generate edits JCodec tool. randomly sample space using 5 different edit types. find that number patches passing unit tests up 75% higher with LLM-based than standard Insert edits. Further, observe found are generally less diverse compared ran local runtime improvements. Although many improving by LLM-enhanced GI, best patch was GI.

Language: Английский

ChatGPT: The End of Online Exam Integrity? DOI Creative Commons
Teo Sušnjak, Timothy R. McIntosh

Education Sciences, Journal Year: 2024, Volume and Issue: 14(6), P. 656 - 656

Published: June 17, 2024

This study addresses the significant challenge posed by use of Large Language Models (LLMs) such as ChatGPT on integrity online examinations, focusing how these models can undermine academic honesty demonstrating their latent and advanced reasoning capabilities. An iterative self-reflective strategy was developed for invoking critical thinking higher-order in LLMs when responding to complex multimodal exam questions involving both visual textual data. The proposed demonstrated evaluated real subject experts performance (GPT-4) with vision estimated an additional dataset 600 text descriptions questions. results indicate that invoke multi-hop capabilities within LLMs, effectively steering them towards correct answers integrating from each modality into final response. Meanwhile, considerable proficiency being able answer across 12 subjects. These findings prior assertions about limitations emphasise need robust security measures proctoring systems more sophisticated mitigate potential misconduct enabled AI technologies.

Language: Английский

Citations

49
The rise and potential of large language model based agents: a survey DOI

Zhiheng Xi,

Wen-Xiang Chen, Xin Hua Guo

et al.

Science China Information Sciences, Journal Year: 2025, Volume and Issue: 68(2)

Published: Jan. 17, 2025

Language: Английский

Citations

9
Enhancing Accuracy in Large Language Models Through Dynamic Real-Time Information Injection DOI Open Access
Qian Ouyang,

Shiyu Wang,

Bing Wang

et al.

Published: Dec. 26, 2023

This study presents a novel approach to enhance Large Language Models (LLMs) like Alpaca by dynamically integrating real-time information. method addresses the issue of content hallucination and data relevancy automatically collecting current from credible sources into model prompts. Experiments show significant improvement in accuracy decrease hallucination, with manageable increase response time. The research underscores potential integration making LLMs more accurate contextually relevant, setting foundation for future advancements dynamic processing AI.

Language: Английский

Citations

32
Generative AI in innovation and marketing processes: A roadmap of research opportunities DOI Creative Commons
Paola Cillo, Gaia Rubera

Journal of the Academy of Marketing Science, Journal Year: 2024, Volume and Issue: unknown

Published: Aug. 26, 2024

Abstract Nowadays, we are witnessing the exponential growth of Generative AI (GenAI), a group models designed to produce new content. This technology is poised revolutionize marketing research and practice. Since literature about GenAI still in its infancy, offer technical overview how trained they Following this, construct roadmap for future on marketing, divided into two main domains. The first domain focuses firms can harness potential throughout innovation process. We begin by discussing changes consumer behavior propose questions at level. then connect these emerging insights with corresponding firm strategies, presenting second set examines likely consequences using analyze: (1) relationship between market-based assets value, (2) skills, preferences, role processes.

Language: Английский

Citations

12
A Generative Artificial Intelligence Using Multilingual Large Language Models for ChatGPT Applications DOI Creative Commons
Nguyen Trung Tuan, Philip Moore,

Dat Ha Vu Thanh

et al.

Applied Sciences, Journal Year: 2024, Volume and Issue: 14(7), P. 3036 - 3036

Published: April 4, 2024

ChatGPT plays significant roles in the third decade of 21st Century. Smart cities applications can be integrated with various fields. This research proposes an approach for developing large language models using generative artificial intelligence suitable small- and medium-sized enterprises limited hardware resources. There are many AI systems operation development. However, technological, human, financial resources required to develop impractical enterprises. In this study, we present a proposed reduce training time computational cost that is designed automate question–response interactions specific domains smart cities. The model utilises BLOOM as its backbone maximum effectiveness We have conducted set experiments on several datasets associated validate model. Experiments English Vietnamese languages been combined low-rank adaptation cost. comparative experimental testing, outperformed ‘Phoenix’ multilingual chatbot by achieving 92% performance compared ‘ChatGPT’ benchmark.

Language: Английский

Citations

10
(Ir)rationality and cognitive biases in large language models DOI Creative Commons
Olivia Macmillan-Scott, Mirco Musolesi

Royal Society Open Science, Journal Year: 2024, Volume and Issue: 11(6)

Published: June 1, 2024

Do large language models (LLMs) display rational reasoning? LLMs have been shown to contain human biases due the data they trained on; whether this is reflected in reasoning remains less clear. In paper, we answer question by evaluating seven using tasks from cognitive psychology literature. We find that, like humans, irrationality these tasks. However, way displayed does not reflect that humans. When incorrect answers are given tasks, often ways differ human-like biases. On top of this, reveal an additional layer significant inconsistency responses. Aside experimental results, paper seeks make a methodological contribution showing how can assess and compare different capabilities types models, case with respect reasoning.

Language: Английский

Citations

9
Inductive Reasoning of Vocational Technical Instructor Students DOI
Katalin Kanczné Nagy, Péter Tóth

Lecture notes in networks and systems, Journal Year: 2025, Volume and Issue: unknown, P. 241 - 252

Published: Jan. 1, 2025

Language: Английский

Citations

0
Making sense of transformer success DOI Creative Commons
Nicola Angius, Pietro Perconti, Alessio Plebe

et al.

Frontiers in Artificial Intelligence, Journal Year: 2025, Volume and Issue: 8

Published: April 1, 2025

This article provides an epistemological analysis of current attempts explaining how the relatively simple algorithmic components neural language models (NLMs) provide them with genuine linguistic competence. After introducing Transformer architecture, at basis most NLMs, paper firstly emphasizes central question in philosophy AI has been shifted from "can machines think?", as originally put by Alan Turing, to "how can pointing explanatory gap for NLMs. Subsequently, existing strategies functioning NLMs are analyzed argue that they, however debated, do not differ used cognitive science explain intelligent behaviors humans. In particular, available experimental studies turned test theory mind, discourse entity tracking, and property induction examined under light functional science; so-called copying algorithm head phenomenon a shown mechanist explanation in-context learning; finally, pioneering use predict brain activation patterns when processing here involve what we call co-simulation, which NLM simulate understand each other.

Language: Английский

Citations

0
Narrative coherence in neural language models DOI Creative Commons
Alessandro Acciai,

Lucia Guerrisi,

Pietro Perconti

et al.

Frontiers in Psychology, Journal Year: 2025, Volume and Issue: 16

Published: April 1, 2025

Neural language models, although at first approximation they may be simply described as predictors of the next token in a given sequence, surprisingly exhibit linguistic behaviors akin to human ones. This suggests existence an underlying sophisticated cognitive system production. intriguing circumstance has inspired adoption psychological theories investigative tools and present research falls within this line inquiry. What we aim establish is potential core coherent integration production, metaphorically parallel speaker's personal identity. To investigate this, employed well-established theory on narrative coherence autobiographical stories. offers theoretical advantage strong correlation between high integrative level knowledge system. It also provides empirical methodologies for quantifying its characteristic dimensions through analysis texts. The same methodology was applied 2010 stories generated by GPT-3.5 equal number from GPT-4, elicited asking models assume roles that included variety variables such gender, mood, age. large ensures adequate sampling stochastic nature made possible thanks automated evaluation procedure. We initially asked generate 192 stories, which were then analyzed team professional psychologists. Based sample, constructed training set fine-tuning automatic evaluator. Our results 4020 overall show fully with data subjects, slightly higher values case GPT-4. These suggest unification comparable self beings.

Language: Английский

Citations

0
A thematic analysis of what Australians state would change their minds on climate change DOI Creative Commons
Amy S G Lee, Kelly Kirkland, Samantha K. Stanley

et al.

Scientific Reports, Journal Year: 2025, Volume and Issue: 15(1)

Published: April 22, 2025

Language: Английский

Citations

0