Skip to main content

Introduction

Background​

The Walnut Plan aspires to be an open pioneer in the research of complex reasoning and deep thinking capabilities in Artificial Intelligence. Our mission is to explore, develop, and share innovative technologies that significantly enhance the complex reasoning abilities of large language models, while advocating for the principles of open science and knowledge sharing. Through relentless efforts, we aim to drive the evolution of AI from simple information processing tools to intelligent systems capable of deep thinking at the level of "Newton" and "Einstein".

Mission​

Guided by this grand vision, the Walnut Plan is committed to reshaping the paradigm of AI research. We believe that only through open, transparent, and collaborative approaches can we truly promote the democratization of AI technology for the benefit of all humanity. Our goals extend beyond technological breakthroughs to fostering a more open and responsible AI research ecosystem.

In the long term, the Walnut Plan will dedicate itself to developing a new generation of AI systems capable of assisting and even independently conducting scientific discoveries. We expect to accelerate the accumulation of human knowledge and innovation processes through the popularization of AI capabilities. At the same time, we are deeply aware of our significant responsibility and will always prioritize ensuring the safety and ethical development of AI, exploring new ways for AI and human intelligence to develop in synergy.

By establishing an interdisciplinary and inter-organizational research network, we aim to bring together top global talent to collectively tackle the frontier challenges in the AI field. We believe that only through collective wisdom and effort can we truly achieve leap-forward development in AI technology and make substantial contributions to human technological progress and social development.

The Walnut Plan is not just a research initiative, but a movement to drive the transformation of AI technology and scientific research culture. We invite all like-minded researchers, developers, and thinkers to join our ranks, to collectively write a new chapter in artificial intelligence, and to contribute to creating a more intelligent, open, and better future.

Modern AI Research​

Modern AI research often involves prolonged, collaborative efforts that extend far beyond the traditional scope of individual, short-term projects. These endeavors frequently require large teams, extensive computational resources, and iterative experimentation over the course of months or even years. As a result, the traditional approach of publishing findings only at the end of the research cycle is no longer sufficient to meet the demands of the field.

Moreover, the fast pace of AI advancements and the need for reproducibility call for more open and continuous communication of research progress. Researchers are increasingly recognizing the value of sharing not just the final results but also the tools, datasets, methodologies, and even setbacks encountered along the way. This approach not only promotes transparency but also helps the broader research community to learn from challenges, adapt methodologies, and contribute to a collective knowledge base that accelerates progress across the field.

We represents a pioneering effort to reimagine the process of conducting and communicating AI research. By providing a comprehensive, real-time account of our journey to replicate the groundbreaking O1 model, we aim to:

  • Addressing the Challenges of Modern AI Research: The progress report tackles the growing complexity of AI research by offering real-time updates, enhancing transparency, and maintaining researcher motivation throughout prolonged, team-based projects.

  • Fostering Open Science and Collective Advancement: By sharing tools, datasets, methodologies, and setbacks, the report promotes knowledge sharing and accelerates progress across the research community, aligning with the principles of open science.

  • Laying the Foundation for AI in Scientific Discovery: Detailed documentation of the research process, including both successes and failures, provides a valuable dataset for training AI systems capable of understanding and contributing to scientific discovery, beyond mere outcomes.

  • Promoting Responsible AI Development: The report encourages transparency in AI research by documenting decision-making processes and ethical considerations, fostering trust and accountability in the development of AI technologies.