Search papers, labs, and topics across Lattice.
Affiliation:
2
0
3
1
Serving costs of memory systems can deviate by up to 69% from predictions based on conversation length, revealing hidden complexities in agentic memory performance.
Fact-based memory systems can become more cost-effective than long-context LLMs for persistent agents after just ten interaction turns when dealing with 100k token contexts, challenging the assumption that simply scaling context length is always the best approach.
