AdaExploreBench: Benchmarking LLM Agents for Proactive Exploration and Test-time Adaptation in Evolving Environments

A new benchmark, AdaExploreBench, is designed to test whether large language model (LLM) agents can proactively explore unfamiliar, evolving environments and adapt at test time, addressing a gap the authors say existing benchmarks rarely cover. The benchmark targets agents' ability to discover how an environment works and adjust to its dynamics rather than relying on static tasks.

Large Language Model LLM agents are increasingly deployed in unfamiliar and evolving environments, requiring them to proactively explore, understand, and adapt to environmental dynamics. However, existing benchmarks rarely test whether agents can proactively discover how an environment works and a