Related Concept Videos
Language and Cognition
Higher Mental Functions of the Brain: Language
Language formation and comprehension take place in the dominant hemisphere. The dominant hemisphere is responsible for understanding the meaning of spoken, written, or sign language, as well as the ability to communicate. For most people, the left hemisphere is the dominant one. The right hemisphere, then, gives tone and emotional context to the...
Language Development
The critical period for language acquisition suggests that the ability to acquire language is at its peak early in life. As people age, this proficiency decreases. Language development begins very...
Self-Evaluation: Self-Enhancement and Self-Verification
Elaborative Rehearsals
The effectiveness of...
Self-Schemas
You might also read
Related Articles
Articles linked to this work by shared authors, journal, and citation graph.
Temporal and Behaviour-Aware Multimodal Modelling for Hour-Ahead Hypoglycaemia Prediction During Ramadan Fasting in Type 1 Diabetes.
Sustainable E-Health: Energy-Efficient Tiny AI for Epileptic Seizure Detection via EEG.
Graphical Insight: Revolutionizing Seizure Detection with EEG Representation.
Using Unstated Cases to Correct for COVID-19 Pandemic Outbreak and Its Impact on Easing the Intervention for Qatar.
Highly accurate prediction of flammability limits of chemical compounds using novel integrated hybrid models.
Peripheral B-cell receptor repertoire predicts immune-related adverse events following immune checkpoint inhibitor therapy in advanced renal cell carcinoma.
Effects of black soldier fly (Hermetia illucens L.) larvae zoocompost on the mineral element content of blue honeysuckle berries.
Investigation on absorption refrigeration performance of R1243zf with imidazolium ionic liquid as the working pairs.
DeepTriage-CN: integrating clinical text with vital signs for emergency department admission prediction in an aging population.
Gold nanoparticles as dual-action antiviral agents: disruption of SARS-CoV-2 viral envelopes and RNA integrity.
Comparison of capillary microsampling and venous blood for multi-pathogen serosurveillance.
Related Experiment Video
Updated: May 20, 2025

Augmenting Large Language Models via Vector Embeddings to Improve Domain-Specific Responsiveness
Published on: December 6, 2024
Efficient self-attention with smart pruning for sustainable large language models.
Samir Brahim Belhaouari1, Insaf Kraidia2
1Division of Information and Computing Technology, College of Science and Engineering, Hamad Bin Khalifa University, Ar-Rayyan, Qatar. sbelhaouari@hbku.edu.qa.
This study introduces a novel compression method for Large Language Models (LLMs), significantly reducing their size and environmental impact. The technique achieves substantial model compression while maintaining high accuracy, paving the way for more sustainable AI.
Area of Science:
- Artificial Intelligence
- Machine Learning
- Natural Language Processing
Background:
- Large Language Models (LLMs) offer advanced multitasking capabilities but suffer from high computational demands, leading to significant environmental concerns regarding energy and water consumption.
- The internal transformer layers are primary contributors to the computational complexity and resource intensity of LLMs.
Purpose of the Study:
- To propose and evaluate an innovative compression approach for reducing the size and computational footprint of Large Language Models (LLMs).
- To address the environmental impact associated with the high energy and water consumption of LLMs through efficient model compression.
Main Methods:
- The proposed method combines mathematical and structural techniques for model compression, focusing on transformer layers.
- Forward Propagation Pruning (FPP) is employed to compress embedding and feed-forward layers using weight freezing and zeroing.
- Weight Matrix Folding, incorporating Identical Row Compression (IRC) and Diagonal Weight Compression (DWC), is used to prune self-attention layer matrices.
Main Results:
- The compression approach achieved a 99% compression of transformer layers and 70% for linear layers, leading to an overall model compression of approximately 70%.
- Model accuracy was maintained at nearly original levels post-compression.
- Moderate compression rates (20-40%) demonstrated stable or improved model performance, alongside significant reductions in memory usage and computational demands.
Conclusions:
- The developed compression technique effectively reduces LLM size and resource requirements, making them more energy-efficient.
- This approach offers a viable path towards more sustainable AI development and deployment.
- Optimized LLMs can achieve performance benefits while minimizing environmental impact.

