MemBench: Towards More Comprehensive Evaluation on the Memory of LLM-based Agents

Open in new window