NirDiamant/Agent_Memory_Techniques
A set of 30 runnable Jupyter notebooks teaching memory implementation patterns for LLM agents.

Velocity · 7d
+15
★ / day
Trend
→steady
star history
This repository provides educational notebooks on agent memory techniques for LLMs. It covers conversation buffers, vector stores, knowledge graphs, episodic and semantic memory, and working memory patterns. The notebooks implement systems including MemGPT, Mem0, Letta, Zep, and Graphiti, with benchmarks like LoCoMo. Each notebook is runnable and targets developers building production-grade LLM agents.