Latest Articles for Quantization

HoSZp allows efficient computations on compressed scientific data, improving analysis workflows.

Learn how language models on devices improve speed and privacy.

A new method makes using large language models on mobile devices more efficient.

This article explores zero-shot quantization and its applications in infrared imaging.

New strategies simplify decoding of LDPC codes for faster communications.

LLaMA3-70B faces unique issues with 8-bit quantization affecting its performance.

Discover efficient methods for fine-tuning large language models using Gaussian noise.

New methods enable non-invasive blood pressure monitoring through wearable devices.

An innovative approach to compress advanced models efficiently without losing performance.

New methods improve image generation efficiency on limited devices.

Exploring key concepts and implications of the Stern-Gerlach experiment in quantum physics.

A new method adapts to input signals, improving quantization accuracy.

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.