Breaking the 100M Token Limit: EverMind's MSA Architecture Achieves Efficient End-to-End Long-Term Memory for LLMs

The research introduces a novel memory architecture called MSA (Memory Sparse Attention). Through a combination of the Memory Sparse Attention mechanism, Document-wise RoPE for extreme context ...
IEEE

Can Large Language Models Perform Retrieval-Augmented Generation as Multi-Hop Reasoning Over Knowledge Graphs?

Abstract: Retrieval-Augmented Generation (RAG) enhances Large Language Models (LLMs) by grounding their outputs in external knowledge. However, conventional chunk-based retrieval methods are limited ...

KIOXIA Achieves 4.8 Billion High-Dimensional Vector Search Database on a Single Server, with 7.8x Index Build Time Acceleration via GPUs

Kioxia America, Inc. today announced the successful demonstration of high-dimensional vector search scaling to 4.8 billion vectors on a single server using its open-source KIOXIA AiSAQ(TM) approximate ...
InfoQ

DoorDash Builds DashCLIP to Align Images, Text, and Queries for Semantic Search Using 32M Labels

DoorDash has launched a multimodal machine learning system that aligns product images, text, and user queries in a shared ...
IEEE

Bauhaus: Restructuring Vector Database for LLM Retrieval on CXL-Based Tiered Memory

Abstract: Retrieval-augmented generation pipelines store large volumes of embedding vectors in vector databases for semantic search. In Compute Express Link (CXL)-based tiered memory systems, ...