AI真能像人类一样做研究吗？新框架DeepResearchEval给出验证

The DeepResearchEval framework presents a groundbreaking approach to evaluating AI's capability in conducting human-like research. This automated system constructs complex research tasks and assesses AI agents' performance across multiple dimensions, including problem formulation, literature synthesis, hypothesis generation, and experimental design. Developed in response to the rapid advancement of large language models (LLMs), the framework addresses ongoing debates about AI's true research potential. Initial evaluations demonstrate AI's competence in structured research tasks while revealing limitations in cross-domain innovation and intuitive reasoning. The study establishes important benchmarks for AI-assisted research and knowledge discovery, with implications for academic research methodologies and AI development roadmaps.