IEEE

Self-Supervised Pretraining and Quantization for Fault Tolerant Neural Networks: Friend or Foe?

Abstract: Deep neural networks (DNNs) are increasingly being applied in critical domains such as healthcare and autonomous driving. However, their predictive capabilities can degrade in the presence ...
IEEE

Oiso: Outlier-Isolated Data Format for Low-Bit Large Language Model Quantization

Abstract: The scale of large language models (LLMs) has steadily increased over time, leading to enhanced performance in multimodal understanding and complex reasoning, but with significant execution ...
GitHub

Python implementation of the TurboQuant and QJL vector quantization algorithms.

turboquant-py implements the TurboQuant and QJL vector quantization algorithms from Google Research (ICLR 2026 / AISTATS 2026). It compresses high-dimensional floating-point vectors to 1-4 bits per ...
GitHub

generate_RD_curves_alexnet.py

# e.g., python generate_RD_curves_vgg16.py 3 1 5 -> run GPU3 to do the statistics for the 5th kernal in layer 1. # the path of checkpoint file of pre-trained VGG16 ...
VentureBeat

Google's new TurboQuant algorithm speeds up AI memory 8x, cutting costs by 50% or more

As Large Language Models (LLMs) expand their context windows to process massive documents and intricate conversations, they encounter a brutal hardware reality known as the "Key-Value (KV) cache ...
Ars Technica

Google’s TurboQuant AI-compression algorithm can reduce LLM memory usage by 6x

Even if you don’t know much about the inner workings of generative AI models, you probably know they need a lot of memory. Hence, it is currently almost impossible to buy a measly stick of RAM without ...