If LLMs are stateless, how can they remember what the user told them in the past? Let’s have a look at the most common implementations of LLM memory, from simple chat history management to RAG-based approaches.
https://www.zansara.dev/posts/2026-02-04-how-does-llm-memory-work/
