Article · AI translation
IT Home reported on October 15 that the Shanghai Artificial Intelligence Laboratory announced today that the 3-dimensional Ghys conjecture, a world-class problem that has challenged mathematicians for more than 30 years, has been solved.
A team led by Xu Disheng, a visiting scholar at the Shanghai Artificial Intelligence Laboratory and a professor at Great Bay University, used the Intern-S2 scientific multimodal large model and a technical approach integrating general and specialized capabilities to complete a proof of the 3-dimensional Ghys conjecture through human–AI collaboration.
A team led by Peking University professor Dong Bin and academician Liu Ruochuan used the SHEAF framework they developed to formally verify the proof, confirming the correctness of the work. This marks the first time Chinese scientists have used a Chinese large model to solve a world-class mathematical problem.
The Ghys conjecture was proposed in 1995 by French mathematician Étienne Ghys. It conjectures that every holomorphic Anosov system on a compact complex manifold is biholomorphically conjugate to a holomorphic affine automorphism on some complex infra-nilmanifold. Both theoretically difficult and significant as a landmark problem in the discipline, it is widely recognized by experts in dynamical systems as one of the field’s important long-standing open problems and serves as a touchstone for mathematical reasoning ability.
The related paper has been published on the preprint platform arXiv, and its formalization code has also been made public on GitHub. IT Home provides the link below:
https://github.com/frenzymath/HolomorphicAnosov
The Shanghai Artificial Intelligence Laboratory said the study is also significant because it validates a new research model. Traditional mathematical research relies on prolonged exploration by individual geniuses, while this effort systematically incorporated expert intuition into a machine-solving loop, creating a cycle of “systematic exploration — expert judgment — accumulation of results — feedback and revision.” This model can be replicated and applied to more areas of basic research where “the search space is vast and experts are scarce,” enabling more original breakthroughs “from 0 to 1.”
Next, drawing on the complete capability loop established in this research, the team will package it as a reusable mathematical verification module for the Intern Duanyan scientific discovery platform, so it can be used directly in subsequent research on scientific problems. At the same time, starting with the breakthrough on the 3-dimensional Ghys conjecture, the Shanghai Artificial Intelligence Laboratory will extend the reasoning capabilities it has developed to other disciplines with related mathematical structures. After achieving new breakthroughs, it will package the new reasoning capabilities back into the Duanyan platform for reuse and combine them with more disciplines to seek the next round of breakthroughs, creating a self-evolving cycle of “breakthrough — packaging — reuse — further breakthrough.”
Advertising notice: External links in this article, including but not limited to hyperlinks, QR codes and access codes, are provided to convey more information and save readers time in selecting information. The results are for reference only. All IT Home articles containing external links include this notice.