Create

Sign in to ReadmeX

or

How we handle your data: Privacy Policy

ReadmeX

Your AI reading companion

What would you like to catch up on?

Pick a briefing, or ask your own question.

Your community and people briefings stay personal to you.

Story

Memorilla releases code for compact semantic memory in LLMs

AI summary

The Memorilla developers released Apache 2.0-licensed code, checkpoints, and evaluation recipes for condensing document collections into a fixed set of query-dependent memory tokens for frozen LLMs. They report that 16 memory tokens achieved 84.33% FactKG accuracy versus 80.43% for full-document prompting while using 108 times fewer tokens, and raised average TextWorld agent success from 18.6% to 24.2%. The repository also includes reinforcement-learning workflows based on TextWorld and GRPO.

In-depth insight

Memorilla provides an implementable approach to giving frozen LLMs compact access to large document collections, with reported token-efficiency and benchmark gains.

🦍 Memorilla: Latent Semantic Memory for LLMs

link

https://memory.stanford.edu


 Memorilla: Latent Semantic Memory for LLMs

This repository provides code for training and evaluating Memorilla, a memory system that gathers information from an entire document collection into a fixed number of query-dependent memory tokens for a frozen LLM. The code is based on our paper Memorilla: Latent Semantic Memory for LLMs.

#Memorilla#Qwen3-8B#Qwen3-Embedding-4B#Mem0#MemoRAG#FactKG#TextWorld#SkyRL-Memorilla

Original · 6 min readRead the original
Group chat

No comments yet. Start the conversation.