RESEARCH · RESEARCH · #2177
QEMem: reframing LLM agent memory as query-specific evidence set construction
This arXiv paper (arXiv:2610.22668v1) identifies a relevance–sufficiency gap in memory use for LLM agents and introduces QEMem, a framework that builds query-specific evidence sets from existing memory stores. QEMem combines query-guided discovery, source-grounded reification into structured ViewCards, and sufficiency-guided refinement, achieving 91.43% LLM-judge accuracy on LoCoMo with Claude Opus 4.5 and outperforming MemR^3 and NEMORI under a shared memory store.
KEY POINTS
- This arXiv paper (arXiv:2610.22668v1) identifies a relevance–sufficiency gap in memory use for LLM agents and introduces QEMem, a framework that builds query-specific evidence sets from existing memory stores.
- QEMem combines query-guided discovery, source-grounded reification into structured ViewCards, and sufficiency-guided refinement, achieving 91.43% LLM-judge accuracy on LoCoMo with Claude Opus 4.5 and outperforming MemR^3 and NEMORI under a shared memory store.
- Addressing the relevance–sufficiency gap can let agents aggregate fragmented or abstracted past records into coherent evidence, improving multi-session reasoning and retrieval accuracy.
WHY IT MATTERS
Addressing the relevance–sufficiency gap can let agents aggregate fragmented or abstracted past records into coherent evidence, improving multi-session reasoning and retrieval accuracy.