BIOENGINEER

Neural Networks Trained on a 24-Dimensional Lattice Could Shrink AI to 0.76 Bits per Weight

A new theoretical framework trains neural networks natively on the Leech lattice, compressing weights to 0.76 bits per ...
Hackaday

vector quantization

Large language models (LLMs) aren’t actually giant computer brains. Instead, they are massive vector spaces in which the probabilities of tokens occurring in a specific order is encoded. Billions of ...
Business Wire

Elastic Announces Faster Filtered Vector Search with ACORN-1 and Default Better Binary Quantization Compression

SAN FRANCISCO--(BUSINESS WIRE)--Elastic (NYSE: ESTC), the Search AI Company, announced new performance and cost-efficiency breakthroughs with two significant enhancements to its vector search. Users ...
Design-Reuse

Pyramid Vector Quantization and Bit Level Sparsity in Weights for Efficient Neural Networks Inference

This paper discusses three basic blocks for the inference of convolutional neural networks (CNNs). Pyramid Vector Quantization [1] (PVQ) is discussed as an effective quantizer for CNNs weights ...
Help Net Security

Google’s TurboQuant cuts AI memory use without losing accuracy

Large language models carry a persistent scaling problem. As context windows grow, the memory required to store key-value (KV) caches expands proportionally, consuming GPU memory and slowing inference ...
Crypto Briefing

Elastic launches Elasticsearch Vector Database for AI search

Elastic N.V. launched its Elasticsearch Vector Database, a serverless AI search product with 32x memory reduction targeting RAG and AI agent ...

Elastic Introduces Serverless Vector Database: Ship in Minutes, Scale Affordably to Hundreds of Billions of Vectors

Elastic (NYSE: ESTC) today announced Elasticsearch Vector Database, a new serverless offering purpose-built for large-scale vector search and AI applications. Elasticsearch is already one of the most ...