← Latest papers
⚛️ quantum physics

No complete set of mutually unbiased bases in dimension ten contains the Fourier pair or three product bases

This paper employs computer-assisted proofs to demonstrate that no complete set of mutually unbiased bases exists in dimension ten if it contains either the Fourier pair or three product bases, by proving that every triple of product bases is strongly unextendible and that the standard-Fourier pair cannot be extended to a set of four mutually unbiased bases.

Original author: Jinze Yang, Yile Liu, Yangyang Li, Xiu-Hao Deng

Published 2026-11-17
📖 7 min read🧠 Deep dive

Original author: Jinze Yang, Yile Liu, Yangyang Li, Xiu-Hao Deng

Original paper licensed under CC BY 4.0 (http://creativecommons.org/licenses/by/4.0/). ✨ This is an AI-generated explanation of the paper below. It is not written or endorsed by the author. For technical accuracy, refer to the original paper. Read full disclaimer

In the quantum world, information is stored in systems that can exist in many states at once. To read or write this information, scientists use special sets of directions, called bases, which act like different ways of looking at the same object. Imagine trying to describe a spinning top; you could look at it from the front, the side, or from above. In quantum physics, these different viewpoints are "mutually unbiased" if knowing the state of the top in one view gives you absolutely no clue about what you would see in another. The more of these distinct, unbiased views you have, the more powerful and secure your quantum system can be. For decades, mathematicians have wondered how many such views are possible for systems of different sizes. When the size of the system is a simple prime number, the answer is known and complete. But for sizes that are made of two numbers multiplied together, like ten, the answer has remained a stubborn mystery.

The number ten is the smallest size where the full picture is missing. We know there are at least three such views, but we do not know if there are four, or even the maximum possible eleven. This gap has left physicists unable to fully design certain types of quantum computers and secure communication networks. A team of researchers has now taken a massive step forward in solving this puzzle for the number ten. They did not find the complete set of views, but they proved that two very natural and promising ways of building such a set are impossible. By using powerful computers to check millions of possibilities with mathematical certainty, they showed that you cannot build a complete set of quantum views in dimension ten if you try to use three specific types of "product" views, nor can you do it if you start with the standard view and its most famous partner, the Fourier view.

To understand what the researchers did, one must first see the landscape they were exploring. In a system of size ten, which can be thought of as a combination of a two-part system and a five-part system, there are special views called product bases. These are views where the information in the two parts is completely separate, like looking at a red ball and a blue ball independently rather than as a single mixed object. Another famous pair of views involves the standard way of listing states and the Fourier view, which mixes them in a specific, rhythmic pattern. For a long time, it was hoped that a complete set of eleven views might be built by combining these natural ingredients. The researchers set out to test this hope. They asked a simple question: if you take three of these product views, or if you take the standard view and the Fourier view, is it possible to add a fourth view that is unbiased to all of them?

The answer, they proved, is a definitive no. For the product views, the team showed that no matter how you arrange three of them, there is no single direction in the quantum space that remains unbiased to all three. It is as if you have three different maps of a city, and you are looking for a single street that looks completely different on all three maps; they proved that such a street does not exist. This result is significant because it rules out a whole class of potential solutions. If a complete set of views existed, it could not contain three of these product views. This settles a long-standing expectation that had been suspected for the number ten but never proven. The team did this by translating the problem into a set of geometric equations and then using a computer to check every possible arrangement. They broke the problem into thousands of small pieces, checking each one with rigorous mathematical logic that leaves no room for rounding errors or guesswork.

The second part of their work focused on the Fourier pair. They investigated whether the standard view and the Fourier view could be part of a larger group of four unbiased views. First, they counted exactly how many directions are unbiased to both of these starting views. They found there are exactly 3,040 such directions. This number had been guessed by other scientists using numerical methods, but this team provided a proof that the count is exact. They then examined how these 3,040 directions relate to one another. They discovered that each of these directions is unbiased to at most eight of the others. This limitation is the key to the proof. If a fourth view existed, it would need to contain ten directions, all of which are unbiased to the starting pair. But since no single direction in the set of 3,040 is unbiased to more than eight others, it is impossible to find ten directions that are all mutually unbiased. Therefore, no fourth view can exist.

The researchers did not stop at the proof; they also mapped out the structure of these 3,040 directions in great detail. They found that 40 of them are the special product directions mentioned earlier, while the rest belong to different groups based on their symmetry. They verified that even if you try to build a third view using any of these directions, you will always hit a wall that prevents a fourth view from being added. The entire process was a feat of computational mathematics. The team used two different computer programs, written in different languages and running on different machines, to check the same results. They ensured that every step was verified by independent methods, from the initial counting of roots to the final check of the geometric relationships. This double-checking guarantees that the result is not an artifact of a computer glitch or a floating-point error, but a solid mathematical fact.

The implications of this work are clear, even if the final answer to the ten-dimensional puzzle remains incomplete. The researchers have removed two of the most likely candidates for a complete set of views. They have shown that the path to a complete set does not go through three product views, nor does it go through the standard and Fourier pair. This narrows the search significantly for anyone trying to find the complete set. It suggests that if a complete set of eleven views exists in dimension ten, it must be built from a very different kind of structure, one that does not rely on these familiar patterns. The work also highlights the power of modern computational proofs. By combining deep mathematical theory with rigorous computer verification, scientists can now tackle problems that were previously thought to be too complex for human calculation alone.

The study leaves the door open for the possibility that a complete set of views exists, but it closes the door on the most obvious ways to find it. The researchers have mapped the terrain and shown where the dead ends are. They have proven that the landscape of dimension ten is more restrictive than previously thought. While the question of whether eleven views exist remains unanswered, the team has established that any such set must be exotic and unexpected. Their work stands as a testament to the precision of modern science, where a combination of human insight and machine power can definitively rule out entire classes of possibilities, bringing us one step closer to understanding the fundamental limits of quantum information.

Drowning in papers in your field?

Get daily digests of the most novel papers matching your research keywords — with technical summaries, in your language.

Try Digest →