⚛️ quantum physics

Exploring Bell Nonlocality with Extremal Non-Signaling Boxes

This paper presents a comprehensive classification of extremal non-signaling (ENS) boxes in arbitrary bipartite Bell scenarios and leverages them to resolve foundational questions, including demonstrating their violation of exclusivity and Specker's principles, decomposing the magic square correlation, and identifying communication thresholds for their simulation.

Original authors: Emmanuel Zambrini Cruzeiro, Junior R. Gonzales-Ureta, Raman Choudhary, Hugo Abreu, Adán Cabello, Sébastien Designolle

Published 2026-10-07

📖 5 min read🧠 Deep dive

CC BY 4.0

Original authors: Emmanuel Zambrini Cruzeiro, Junior R. Gonzales-Ureta, Raman Choudhary, Hugo Abreu, Adán Cabello, Sébastien Designolle

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 authors. For technical accuracy, refer to the original paper. Read full disclaimer

In the strange world of quantum physics, particles can become linked in ways that seem to defy common sense. When two such particles are measured, the result of one measurement can instantly influence the other, no matter how far apart they are. This phenomenon, known as nonlocality, is a verified fact of nature. However, physicists have long wondered about the absolute limits of these connections. To explore this, they imagine a theoretical machine called a "box." This box takes inputs from two people, Alice and Bob, and produces outputs that are perfectly correlated. While real quantum machines can produce very strong correlations, there is a theoretical ceiling. Some imagined boxes could produce correlations so strong that they would break the rules of information theory, allowing for impossible feats like instant communication or solving complex problems with zero effort. These theoretical extremes are called extremal non-signaling boxes. They represent the outermost boundary of what is mathematically possible without allowing faster-than-light communication. Understanding where the real quantum world sits in relation to these impossible extremes helps scientists figure out why nature chose the rules it did.

A team of researchers has now mapped out these impossible extremes in a much wider variety of situations than ever before. For years, scientists mostly studied the simplest possible scenario involving two people with two choices each. In that narrow case, there is only one type of extreme box, famously known as the Popescu-Rohrlich box. The new work expands this view to many more complex scenarios, where Alice and Bob have more choices and more possible outcomes. By combining clever mathematical constructions with powerful computer algorithms, the team generated a complete list of these extreme boxes for several new, unexplored situations. They found thousands of distinct types of these boxes, creating a detailed catalog that had never existed. This list serves as a new map, allowing researchers to see exactly how the quantum world is bounded by these theoretical limits.

One of the first things the team did with this new map was to break down a famous quantum phenomenon called the "magic square" correlation. This is a perfect quantum strategy that wins a specific game every single time. The researchers showed that this perfect quantum result can be built by mixing together just two of these extreme, impossible boxes. This is a significant discovery because it reveals that even the most perfect quantum behaviors can be understood as a simple combination of these theoretical extremes. It suggests that the strange rules of quantum mechanics might be a specific, limited blend of the broader possibilities allowed by the laws of logic.

The team also tested a fundamental principle called local orthogonality. This principle states that if you have two independent copies of a system, the total probability of certain events happening together cannot exceed a specific limit. For a long time, it was known that the simplest extreme boxes violate this rule when you have two copies of them. The researchers tested every single type of extreme box they found in their new, larger scenarios. In every case, they found that two copies of the box were enough to break the local orthogonality rule. This supports a strong suspicion that this violation is a universal feature of these extreme boxes in two-party scenarios. It implies that nature strictly forbids these extreme correlations, not just because of how they behave alone, but because they become even more impossible when duplicated.

Finally, the researchers investigated how much information would be needed to simulate these impossible boxes using ordinary classical communication. They asked a simple question: if Alice and Bob were allowed to send a single message to each other, how many different values would that message need to have to perfectly mimic the behavior of these extreme boxes? They found that for many of the boxes they studied, a single bit of information (a simple yes or no) was not enough. In fact, for a large number of cases, even a message with five different possible values was insufficient. They identified specific scenarios where the communication cost is so high that it requires a message with up to five levels of information. This work provides a clear measure of the "distance" between the quantum world and these theoretical extremes, showing that bridging the gap requires more communication than previously thought possible.

The study does not claim to have found a new physical particle or a new law of nature. Instead, it provides a rigorous, computational framework for understanding the landscape of possibilities. By cataloging these extreme boxes and testing their properties, the researchers have turned abstract questions about the limits of quantum mechanics into concrete, solvable problems. They have shown that the tools needed to understand quantum nonlocality are more diverse and complex than the single famous example used for decades. This new catalog of extreme boxes offers a fresh perspective, suggesting that the rules of our universe are defined by a delicate balance between what is logically possible and what is physically realizable.

Problem Statement
Extremal non-signaling (ENS) boxes are vertices of the non-signaling (NS) polytope in Bell scenarios. While neither quantum theory nor theories of ideal measurements permit nonlocal ENS boxes (i.e., vertices in NS that are not in the local polytope $L$ ), these boxes are fundamental to understanding the boundaries of quantum correlations. They serve as upper bounds for the quantum set and are essential for addressing problems in Bell nonlocality, such as characterizing the resources required to simulate quantum correlations or testing operational principles like the exclusivity principle (local orthogonality).

Despite their importance, a comprehensive understanding of ENS boxes is limited. While complete lists exist for specific scenarios like $(d, d, 2, 2)$ , $(2, 2, d, d)$ , and $(2, 2, 2, 2, 2, 2)$ , a general characterization of ENS vertices in arbitrary bipartite Bell scenarios is lacking. Furthermore, foundational questions remain regarding the minimal decomposition of perfect quantum correlations (such as the magic square game) into ENS boxes, the extent to which nonlocal ENS boxes violate local orthogonality (LO) principles, and the classical communication costs required to simulate them.

Methodology
The authors employ a combination of explicit analytical constructions and computational polytope methods to explore Bell nonlocality:

Construction of Candidate Boxes: To generate candidate nonlocal ENS boxes in arbitrary bipartite scenarios $(X, Y, A, B)$ , the authors combine structured ansätze based on the constructions of Jones and Masanes [14] and Barrett et al. [15]. These candidates are defined by block matrices where specific sub-blocks follow circulant or permutation structures.
Vertex Enumeration and Certification: The software package PANDA is used to verify the extremality of these candidates within the non-signaling polytope. The workflow involves:
- Generating candidates using the structured block ansatz.
- Testing candidates for extremality.
- Feeding certified vertices and symmetry generators into PANDA to enumerate vertex classes.
- Determining if the resulting list is complete or conjectural based on whether the enumeration process terminates with a closed set of inequalities.
Decomposition Analysis: The authors derive minimal decompositions of the magic square correlations into ENS boxes by solving for vectors that satisfy positivity, normalization, and non-signaling constraints, aiming to saturate constraints to find extremal points.
Principle Violation Testing: To test the exclusivity principle (Local Orthogonality, LO), the authors construct joint exclusivity graphs for multiple copies of a box (specifically $LO_2$ , involving two copies). They search for cliques in these graphs where the sum of joint probabilities exceeds one.
Communication Complexity Analysis: Using an adaptation of the Frank–Wolfe technique, the authors determine the minimum classical communication cost (specifically one-way messages of size $d$ , or "dits") required to simulate specific ENS boxes. They compare the performance of these boxes against the polytope of local hidden variable models supplemented with classical communication ( $LHV + \text{dit}$ ).

Key Contributions and Results

Minimal Decomposition of Magic Square Correlations: The authors provide a minimal decomposition of the magic square correlations (a perfect quantum correlation in the $(3,3,4,4)$ scenario) into exactly two ENS boxes. They explicitly construct these two boxes, $\vec{p}_1$ and $\vec{p}_2$ , proving they are extremal and that the magic square correlation is their midpoint. They also provide a decomposition into eight ENS vertices involving PR-box-like structures.
Complete and Partial Classifications: Using PANDA and the candidate generation method, the authors obtain complete or partial classifications of ENS boxes for several previously unexplored scenarios, including $(2,3,3,2)$ , $(3,3,3,2)$ , $(2,3,3,3)$ , and $(5,4,2,2)$ . For the $(3,3,3,3)$ scenario, they identify 8,747 classes of boxes, conjecturing this to be the complete list. A database containing these lists is provided.
Violation of Local Orthogonality ( $LO_2$ ): The study finds that for every nonlocal ENS box tested, two copies of the box violate the $LO_2$ inequality. This supports the conjecture that in bipartite Bell scenarios, two copies of any nonlocal ENS box are sufficient to violate local orthogonality, distinguishing them from certain tripartite or Kochen–Specker scenarios where single copies or higher-order copies behave differently.
Communication Complexity Bounds: The authors identify minimal scenarios where a single dit of classical communication (with $d \leq 5$ ) is insufficient to simulate specific ENS boxes. They demonstrate that while some boxes (like the $(3,3,3,3)$ box from Ref. [14]) can be simulated with one bit, many others in the $(3,3,3,3)$ scenario require a trit. They provide specific examples of boxes requiring up to a quint ( $d=5$ ) for simulation, establishing upper bounds on the number of PR boxes needed to simulate these nonlocal behaviors.

Significance and Claims
The paper claims that the "ENS-box approach" transforms qualitative questions about Bell nonlocality into concrete, solvable resource-theoretic problems. By moving beyond the paradigm of PR boxes, the authors demonstrate that:

Access to general ENS vertices allows for the systematic probing of Bell nonlocality through resource theories.
These boxes yield new constraints via the exclusivity/LO principle and reveal minimal classical communication simulation costs.
They provide minimal decompositions of canonical perfect quantum correlations.

The authors position their work not merely as an enumeration exercise but as a methodological advancement that opens new avenues for research, including the generation of new Bell inequalities, the analysis of post-quantum correlations, and applications in quantum communication and security proofs under no-signaling adversaries. They explicitly state that their results show the ENS-box approach leads to new results and opens new avenues for research.

More like this