September 12, 2025ElectronicsOpen Access

A Survey of Large Language Models: Evolution, Architectures, Adaptation, Benchmarking, Applications, Challenges, and Societal Implications

View Full Paper

Ask AI

Bookmark

Authors

Discussion

Survey reveals significant implications of large language models in AI, indicating adaptation and benchmarking issues.

High training and inference costs in large language models raise economic and labor concerns, alongside the necessity for human oversight.
The survey highlights critical open problems like detectability, data contamination, and benchmark diversity in large language models' applications.
Unified multimodal architectures are emerging, capable of handling diverse data, showing potential in various sectors including healthcare and engineering.
Collaborative advancements in adaptation and evaluation are essential to tackle the complex challenges posed by large language models.

Mohammadabadi et al. (2025) studied this question.

View Full Paper

Ask AI

Bookmark