EM-LLM: Human-Inspired Episodic Memory for Infinite Context LLMs
9 jbotz 1 5/10/2025, 7:49:18 AM github.com โ
Comments (1)
MacsHeadroom ยท 12m ago
So, infinite context length by making it compute bound instead of memory bound. Curious how much longer this takes to run and when it makes sense to use vs RAG.