---
format: "aidr-story-markdown/v1"
id: "bc00e1abe23c7a7329aa814fe5150dfaef6f5ba5421ddaf868c92523c290ff65"
canonical_url: "https://aidr.today/bc00e1ab?lang=en"
title: "Memorilla: Latent Semantic Memory for LLMs"
lang: "en"
requested_lang: "en"
available_langs: ["en","vi"]
translation_fallback: null
fallback_fields: []
published_at: "2026-11-06T18:03:33.000Z"
category: "Data"
topics: ["llm","data-engineering"]
source_urls: ["https://arxiv.org/abs/2610.21375","https://news.ycombinator.com/item?id=50395919"]
summary: "Memorilla is a memory system that gathers information from an entire document collection into a fixed number of query-dependent 'memory tokens' for a frozen LLM. Using only 16 memory tokens, Memorilla achieved 84.33% accuracy on FactKG versus 80.43% with full-context inference. Memorilla improved an agent's average success rate from 18.6% to 24.2% in a TextWorld environment."
---

# Memorilla: Latent Semantic Memory for LLMs

> [Open the canonical story](<https://aidr.today/bc00e1ab?lang=en>)

**Published:** 2026-11-06T18:03:33.000Z
**Category:** Data
**Topics:** llm, data\-engineering

## Summary

Memorilla is a memory system that gathers information from an entire document collection into a fixed number of query\-dependent 'memory tokens' for a frozen LLM\. Using only 16 memory tokens, Memorilla achieved 84\.33% accuracy on FactKG versus 80\.43% with full\-context inference\. Memorilla improved an agent's average success rate from 18\.6% to 24\.2% in a TextWorld environment\.

## Sources

- [Story source](<https://arxiv.org/abs/2610.21375>)
- [Discussion](<https://news.ycombinator.com/item?id=50395919>)