Evaluating Cognitive Age Alignment in Interactive AI Agents
For AI researchers, it provides a benchmark to assess how well agentic AI systems match human cognitive development stages, highlighting a critical gap in current AI capabilities.
The paper introduces ChildAgentEval, a psychometrically grounded benchmark for evaluating cognitive age alignment in MLLM-based agents, revealing that state-of-the-art AI agents fail at simple tasks that children can solve.
While agentic AI and its core multimodal large language models (MLLMs) have demonstrated remarkable promise in language and visual reasoning across domains ranging from daily life to advanced scientific research, a profound gap remains between artificial and human intelligence. Despite the integration of powerful tools and advanced MLLMs, state-of-the-art AI agents frequently fail at foundational, seemingly simple tasks that a child can resolve with ease. Inspired by the Wechsler Intelligence Scale for Children (WISC), we introduce ChildAgentEval, the first psychometrically grounded interactive benchmark for evaluating cognitive age alignment in MLLM-based agents. ChildAgentEval systematically compares the reasoning performance of various MLLM-based interactive agents against age-specific human developmental stages, exposing where current agentic AI systems can and cannot simulate age-specific cognitive behavior.