LifelongAgentBench: Evaluating LLM Agents as Lifelong Learners
Fuente:
arXiv
Gespeichert in:
| Hauptverfasser: | , , , , , , , |
|---|---|
| Format: | Preprint |
| Veröffentlicht: |
2025
|
| Schlagworte: | |
| Online-Zugang: | |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| _version_ | 1866918038694002688 |
|---|---|
| author | Zheng, Junhao Cai, Xidi Li, Qiuke Zhang, Duzhen Li, ZhongZhi Zhang, Yingying Song, Le Ma, Qianli |
| author_facet | Zheng, Junhao Cai, Xidi Li, Qiuke Zhang, Duzhen Li, ZhongZhi Zhang, Yingying Song, Le Ma, Qianli |
| contents | Lifelong learning is essential for intelligent agents operating in dynamic environments. Current large language model (LLM)-based agents, however, remain stateless and unable to accumulate or transfer knowledge over time. Existing benchmarks treat agents as static systems and fail to evaluate lifelong learning capabilities. We present LifelongAgentBench, the first unified benchmark designed to systematically assess the lifelong learning ability of LLM agents. It provides skill-grounded, interdependent tasks across three interactive environments, Database, Operating System, and Knowledge Graph, with automatic label verification, reproducibility, and modular extensibility. Extensive experiments reveal that conventional experience replay has limited effectiveness for LLM agents due to irrelevant information and context length constraints. We further introduce a group self-consistency mechanism that significantly improves lifelong learning performance. We hope LifelongAgentBench will advance the development of adaptive, memory-capable LLM agents. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2505_11942 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | LifelongAgentBench: Evaluating LLM Agents as Lifelong Learners Zheng, Junhao Cai, Xidi Li, Qiuke Zhang, Duzhen Li, ZhongZhi Zhang, Yingying Song, Le Ma, Qianli Artificial Intelligence Lifelong learning is essential for intelligent agents operating in dynamic environments. Current large language model (LLM)-based agents, however, remain stateless and unable to accumulate or transfer knowledge over time. Existing benchmarks treat agents as static systems and fail to evaluate lifelong learning capabilities. We present LifelongAgentBench, the first unified benchmark designed to systematically assess the lifelong learning ability of LLM agents. It provides skill-grounded, interdependent tasks across three interactive environments, Database, Operating System, and Knowledge Graph, with automatic label verification, reproducibility, and modular extensibility. Extensive experiments reveal that conventional experience replay has limited effectiveness for LLM agents due to irrelevant information and context length constraints. We further introduce a group self-consistency mechanism that significantly improves lifelong learning performance. We hope LifelongAgentBench will advance the development of adaptive, memory-capable LLM agents. |
| title | LifelongAgentBench: Evaluating LLM Agents as Lifelong Learners |
| topic | Artificial Intelligence |
| url | https://arxiv.org/abs/2505.11942 |