Towards Robust Long-Lived Agents: Benchmarking Experience Lifecycles in Evolving Environments
中文摘要
该研究提出了针对持续性智能体的新基准,旨在评估智能体在不断演变的环境中学习、重用经验及保持知识时效性的能力。
English Summary
This study introduces a new benchmark for long-lived agents, evaluating their ability to learn, reuse experiences, and maintain relevant knowledge within constantly evolving environments.
arXiv:2611.05450v1 Announce Type: new Abstract: As agents become more capable, they are moving beyond one-off tasks toward sustained assistance across successive user goals. Such assistance requires agents to discover how unfamiliar environments work, reuse what they learn across tasks, and keep that experience current as environments evolve. Existing benchmarks increasingly study experience reuse and adaptation, but provide limited insight into which parts of accumulated experience agents revise, retain, or lose. To address this gap, we introduce ReScopeBench, a benchmark for the experience lifecycle across 18 interactive environments. It distinguishes structural experience about how an environment is organized and accessed from rule experience about reusable principles and constraints that guide decisions across tasks. To study how experience is acquired, reused, and revised, we propose an experience-first construction approach. It builds task streams around shared experience and introduces silent environmental changes. We evaluate task performance and track changes in saved experience throughout each stream. Across 20 models, we find that retained experience improves task perfor…