MemTrace: Rethinking Long-Term Memory Evaluation in LLM Agents
A new study, MemTrace, addresses the limitations of current accuracy metrics in assessing long-term memory in large language model agents, which often overlook critical aspects.
Editorial Staff 4 days ago