Towards Fairness-Aware Adversarial Learning DOI
Yanghao Zhang, Tianle Zhang,

Ronghui Mu

et al.

2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), Journal Year: 2024, Volume and Issue: 33, P. 24746 - 24755

Published: June 16, 2024

Language: Английский

A survey of GPT-3 family large language models including ChatGPT and GPT-4 DOI Creative Commons

Katikapalli Subramanyam Kalyan

Natural Language Processing Journal, Journal Year: 2023, Volume and Issue: 6, P. 100048 - 100048

Published: Dec. 19, 2023

Large language models (LLMs) are a special class of pretrained (PLMs) obtained by scaling model size, pretraining corpus and computation. LLMs, because their large size on volumes text data, exhibit abilities which allow them to achieve remarkable performances without any task-specific training in many the natural processing tasks. The era LLMs started with OpenAI's GPT-3 model, popularity has increased exponentially after introduction like ChatGPT GPT4. We refer its successor OpenAI models, including GPT4, as family (GLLMs). With ever-rising GLLMs, especially research community, there is strong need for comprehensive survey summarizes recent progress multiple dimensions can guide community insightful future directions. start paper foundation concepts transformers, transfer learning, self-supervised models. then present brief overview GLLMs discuss various downstream tasks, specific domains languages. also data labelling augmentation robustness effectiveness evaluators, finally, conclude To summarize, this will serve good resource both academic industry people stay updated latest related GLLMs.

Language: Английский

Citations

131
Large AI Models in Health Informatics: Applications, Challenges, and the Future DOI Creative Commons
Jianing Qiu, Lin Li, Jiankai Sun

et al.

IEEE Journal of Biomedical and Health Informatics, Journal Year: 2023, Volume and Issue: 27(12), P. 6074 - 6087

Published: Sept. 22, 2023

Large AI models, or foundation are models recently emerging with massive scales both parameter-wise and data-wise, the magnitudes of which can reach beyond billions. Once pretrained, large demonstrate impressive performance in various downstream tasks. A prime example is ChatGPT, whose capability has compelled people's imagination about far-reaching influence that have their potential to transform different domains our lives. In health informatics, advent brought new paradigms for design methodologies. The scale multi-modal data biomedical domain been ever-expanding especially since community embraced era deep learning, provides ground develop, validate, advance breakthroughs health-related areas. This article presents a comprehensive review from background applications. We identify seven key sectors applicable might substantial influence, including 1) bioinformatics; 2) medical diagnosis; 3) imaging; 4) informatics; 5) education; 6) public health; 7) robotics. examine challenges, followed by critical discussion future directions pitfalls transforming field informatics.

Language: Английский

Citations

122
Evolution and Optimization of Language Model Architectures: From Foundations to Future Directions DOI

Zainab M. AlQenaei

Lecture notes in networks and systems, Journal Year: 2025, Volume and Issue: unknown, P. 233 - 249

Published: Jan. 1, 2025

Language: Английский

Citations

6
Integrating human expertise & automated methods for a dynamic and multi-parametric evaluation of large language models’ feasibility in clinical decision-making DOI Creative Commons
Elena Sblendorio, Vincenzo Dentamaro, Alessio Lo Cascio

et al.

International Journal of Medical Informatics, Journal Year: 2024, Volume and Issue: 188, P. 105501 - 105501

Published: May 26, 2024

Recent enhancements in Large Language Models (LLMs) such as ChatGPT have exponentially increased user adoption. These models are accessible on mobile devices and support multimodal interactions, including conversations, code generation, patient image uploads, broadening their utility providing healthcare professionals with real-time for clinical decision-making. Nevertheless, many authors highlighted serious risks that may arise from the adoption of LLMs, principally related to safety alignment ethical guidelines. To address these challenges, we introduce a novel methodological approach designed assess specific feasibility adopting LLMs within area, focus nursing, evaluating performance thereby directing choice. Emphasizing LLMs' adherence scientific advancements, this prioritizes care personalization, according "Organization Economic Co-operation Development" frameworks responsible AI. Moreover, its dynamic nature is adapt future evolutions LLMs. Through integrating advanced multidisciplinary knowledge, Nursing Informatics, aided by prospective literature review, seven key domains evaluation items were identified follows: State Art Alignment & Safety. Focus, Accuracy Management Prompt Ambiguity. Data Integrity, Security, Ethics Sustainability, accordance OECD Recommendations Responsible Temporal Variability Responses (Consistency) Adaptation standardized terminology Classifications professionals. General Capabilities: Post User Feedback Self-Evolution Capability Organization Chapters. Ability Drive Evolution Healthcare. Nine state art evaluated using methodology oncology nursing decision-making, producing preliminary results. Gemini Advanced, Anthropic Claude 3 4 achieved minimum score Safety domain classification "recommended", being also endorsed across all domains. LLAMA 70B 3.5 classified "usable high caution." Others unusable domain. The identification recommended LLM combined critical, prudent, integrative use, can decision-making processes.

Language: Английский

Citations

12
Safety analysis in the era of large language models: A case study of STPA using ChatGPT DOI Creative Commons
Yi Qi, Xingyu Zhao, Siddartha Khastgir

et al.

Machine Learning with Applications, Journal Year: 2025, Volume and Issue: unknown, P. 100622 - 100622

Published: Jan. 1, 2025

Language: Английский

Citations

1
A Comprehensive Review of Current Trends, Challenges, and Opportunities in Text Data Privacy DOI Creative Commons
Sakib Shahriar,

Rozita Dara,

Rajen Akalu

et al.

Computers & Security, Journal Year: 2025, Volume and Issue: unknown, P. 104358 - 104358

Published: Jan. 1, 2025

Language: Английский

Citations

1
Promoting Cognitive Health in Elder Care with Large Language Model-Powered Socially Assistive Robots DOI
Maria R. Lima, Amy O'Connell,

F.B. Zhou

et al.

Published: April 24, 2025

Language: Английский

Citations

1
A Comprehensive Survey of Attack Techniques, Implementation, and Mitigation Strategies in Large Language Models DOI

Aysan Esmradi,

Daniel Wankit Yip,

Chun Fai Chan

et al.

Communications in computer and information science, Journal Year: 2024, Volume and Issue: unknown, P. 76 - 95

Published: Jan. 1, 2024

Language: Английский

Citations

5
Assessing the Security of GitHub Copilot's Generated Code - A Targeted Replication Study DOI
Vahid Majdinasab,

Michael Joshua Bishop,

Shawn Rasheed

et al.

2022 IEEE International Conference on Software Analysis, Evolution and Reengineering (SANER), Journal Year: 2024, Volume and Issue: unknown, P. 435 - 444

Published: March 12, 2024

Language: Английский

Citations

4
AI-based chatbot interactions and critical thinking skills: an exploratory study DOI
Rosa Angela Fabio, Alessio Plebe, Rossella Suriano

et al.

Current Psychology, Journal Year: 2024, Volume and Issue: unknown

Published: Oct. 12, 2024

Language: Английский

Citations

4