LifelongAgentBench: Evaluating LLM Agents as Lifelong Learners
This addresses the need for adaptive, memory-capable LLM agents in dynamic environments, though it appears incremental as it builds on existing agent frameworks with a new benchmark and improvement mechanism.
The authors tackled the problem that current LLM-based agents lack lifelong learning capabilities by creating LifelongAgentBench, the first unified benchmark to systematically evaluate such abilities across three interactive environments, finding that conventional experience replay has limited effectiveness while their group self-consistency mechanism significantly improves performance.
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.