Causal graph memory for LLMs. Flat token cost, no matter how the session runs

• Posted 2 hours ago by w4mwati
• 3 points