Topic: retrieval

Latest on “retrieval”

The AI news that matters for people who make things — what happened, and how to put it to work in Grafy.

Stanford Open-Sources Memorilla, a 16‑Token Memory Module for Frozen LLMs

Stanford researchers released Memorilla, an open‑source module that compresses a document collection into just 16 memory tokens for frozen Qwen3‑8B.

Try in Grafy Codememory modulefrozen llmopen sourceretrieval

Grafy Research Lab22 h ago5 min read

MIGRO Introduces Decoupled Long‑Term Memory for Evolving Language Models

Researchers unveiled MIGRO, a decoupled long‑term memory system that lets successive language models share exact user histories while using compact storage.

Try in Grafy Chatlong-term memorylanguage models

Grafy Research Labyesterday5 min read

Active Learning Model Cuts LLM Retrieval Energy by 96% for Long Queries

Researchers introduced the LLM Active Approximation Model (LAAM), a hybrid retrieval system that preserves most of the relevance of full LLM reranking.

Try in Readerllmretrieval

Grafy Research Labyesterday4 min read

REWEAVE lets vector indexes learn from queries to boost ANN recall

Researchers unveil REWEAVE, an online method that rewires HNSW graphs using query feedback.

Try in Readervector-searchhnsw

Grafy Research Labyesterday5 min read

AntEmbed Introduces Dual‑Mode Multimodal Embeddings for Faster, More Precise Retrieval

Researchers unveil AntEmbed, a dual‑mode model that merges efficient discriminative embeddings with reasoning‑enhanced generative embeddings.

Try in Readermultimodalembedding

Arxiv DB Recent4 days ago5 min read

MV-IVF Tests Native Indexing for Multi-Vector Retrieval

A new arXiv paper describes MV-IVF, a native IVF-style index for multi-vector retrieval. The source says it targets faster set-based search.

Try in Readermulti-vector-retrievalretrieval

Arxiv AI Recent5 days ago5 min read

LLMs often ignore supplied codebook definitions, study finds

A new arXiv audit shows large language models rarely follow researcher‑provided category definitions.

Try in Readerllmannotation

Towards Data ScienceOct 28, 20265 min read

Your RAG Pipeline Is Missing a 1994 Algorithm, Data Post Says

Towards Data Science says a RAG pipeline may be missing a 1994 algorithm, despite embeddings detecting broad meaning. Here’s why that matters for retrieval.

Try in Readerragembeddings