Abstract
This paper introduces Agent-Based Auto Research, a structured multi-agentframework designed to automate, coordinate, and optimize the full lifecycle ofscientific research. Leveraging the capabilities of large language models(LLMs) and modular agent collaboration, the system spans all major researchphases, including literature review, ideation, methodology planning,experimentation, paper writing, peer review response, and dissemination. Byaddressing issues such as fragmented workflows, uneven methodologicalexpertise, and cognitive overload, the framework offers a systematic andscalable approach to scientific inquiry. Preliminary explorations demonstratethe feasibility and potential of Auto Research as a promising paradigm forself-improving, AI-driven research processes.