Back to Reading Room

LA BIBLIOTHÈQUE · READING NOTE

LongMemEval: Benchmarking Chat Assistants on Long-Term Interactive Memory

Wu et al.

To Read · 想看

AIBenchmarkMemoryPapers

LongMemEval, Long-Term Memory, Temporal Reasoning, Knowledge Update, Benchmark

长期对话记忆 Benchmark,覆盖信息提取、跨 Session 推理、时间推理、知识更新和 Abstention 等能力,用于评估聊天助手是否真的具备长期互动记忆。