Latest Articles for Quantization

Learn how model compression improves efficiency of large language models.

A method to enhance language models' efficiency and performance.

New methods improve neural network performance on limited-resource devices.

RC-FED reduces communication costs while maintaining model quality in federated learning.

This study examines performance and conditions for quantized neural networks under fixed-point arithmetic.

A new algorithm improves coordination among nodes under communication limits.

This article discusses DilateQuant for improving diffusion models' speed and accuracy.

AXE improves model performance while minimizing overflow in accumulator-aware quantization.

A new chatbot assists students with STEM multiple-choice questions.

P4Q combines fine-tuning and quantization for efficient visual-language model performance.

Optimizing DNNs with power-of-two quantization for resource-limited devices.

Innovative methods aim to make large language models more efficient and deployable.

1-bit models show great potential in machine learning efficiency and performance.

Discover how simple tweaks can trick chatbots into unexpected responses.

Learn about quantization and its impact on language models.

Precision impacts the effectiveness and cost of language model training.

Examining how simplifying models affects decision-making clarity and performance.

MicroScopiQ improves AI models' performance while consuming less energy.

QuanCrypt-FL enhances security in Federated Learning using advanced techniques.

A novel method enhances AKNN searches for better speed and accuracy.

Learn how quantization helps optimize large language models for everyday use.

Super weights are key to language model performance and efficiency.

This study examines how large language models can misbehave and be manipulated.

ASER offers a way to enhance quantized language models without losing performance.

Innovative strategies for running advanced AI on mobile devices.

ZipNN compresses AI models efficiently, keeping essential details intact.

Smaller LLMs offer help but have significant quality issues in code generation.

A new method speeds up AI processing without losing accuracy.

Learn how ShiftQuant and L1 normalization improve neural network efficiency.

Keeping AI conversations safe on the go with Llama Guard.

Model compression techniques enable heavy models to run smoothly on smaller devices.

A new method to optimize large language models efficiently.

A study showcasing hybrid architecture for improving SNN performance and energy efficiency.

Research shows how to compress diffusion models while maintaining quality.

Learn about Anda, a new method for managing activation data in LLMs.

Learn how reinforcement learning enhances machine communication and decision-making.

A look into hadrons and their interactions using lattice quantum chromodynamics.

QABBA streamlines time series data analysis for clearer insights.

Discover how AI models can be fast and easy to understand.

Learn how lossless compression is reshaping data storage and processing.