Mnemex FAQs

Question 1

What is Mnemex and how does it enhance AI?

Accepted Answer

Mnemex is a Model Context Protocol (MCP) server that provides AI assistants with human-like memory dynamics. It features a temporal decay algorithm based on the Ebbinghaus forgetting curve, allowing AI to naturally remember and forget information, making interactions more natural and efficient.

Question 2

How does Mnemex manage data privacy and storage?

Accepted Answer

Mnemex prioritizes privacy by storing all data locally on your machine. Short-term memory uses human-readable JSONL files, while long-term memory is stored in Obsidian-optimized Markdown files. This gives users complete control, transparency, and the ability to version control their data.

Question 3

Does Mnemex integrate with Large Language Models (LLMs)?

Accepted Answer

Yes, Mnemex includes a smart prompting system designed for natural LLM integration. This allows AI assistants to automatically save, recall, and reinforce memories during conversation, requiring no explicit memory commands from the user.

Question 4

What is the 'two-layer memory architecture' in Mnemex?

Accepted Answer

Mnemex employs a two-layer system: a Short-Term Memory (STM) for transient context with JSONL storage and temporal decay, and a Long-Term Memory (LTM) for permanent knowledge using Markdown files, often integrated with an Obsidian vault for structured knowledge management.

Question 5

What kind of forgetting dynamics does Mnemex use?

Accepted Answer

Mnemex utilizes a novel temporal decay algorithm inspired by cognitive science. Memories' scores continuously decay over time (recency) but are reinforced by use count (frequency) and an importance parameter. This mimics the Ebbinghaus forgetting curve, where memories fade unless reinforced.

Mnemex

概要

主な機能

ユースケース