The Agent's First Day: Benchmarking Learning, Exploration, and Scheduling in the Workplace Scenarios
This work addresses the need for more realistic evaluation frameworks for AI agents in production-oriented environments, though it is incremental as it builds on existing MLLM research by shifting focus from static to dynamic benchmarks.
The paper tackles the problem of evaluating multi-modal large language models (MLLMs) in dynamic, real-world workplace scenarios, identifying deficiencies in active exploration and continual learning, and introduces a new benchmark environment to assess agent reliability in such settings.
The rapid evolution of Multi-modal Large Language Models (MLLMs) has advanced workflow automation; however, existing research mainly targets performance upper bounds in static environments, overlooking robustness for stochastic real-world deployment. We identify three key challenges: dynamic task scheduling, active exploration under uncertainty, and continuous learning from experience. To bridge this gap, we introduce \method{}, a dynamic evaluation environment that simulates a "trainee" agent continuously exploring a novel setting. Unlike traditional benchmarks, \method{} evaluates agents along three dimensions: (1) context-aware scheduling for streaming tasks with varying priorities; (2) prudent information acquisition to reduce hallucination via active exploration; and (3) continuous evolution by distilling generalized strategies from rule-based, dynamically generated tasks. Experiments show that cutting-edge agents have significant deficiencies in dynamic environments, especially in active exploration and continual learning. Our work establishes a framework for assessing agent reliability, shifting evaluation from static tests to realistic, production-oriented scenarios. Our codes are available at https://github.com/KnowledgeXLab/EvoEnv