Unconditional and exponentially large violation of classicality

arXiv:2511.11008 · quant-ph · Submitted 2025-11-14 · Read on arXiv

Listen

Radio episode about this paper

Transcript

Introduction to the show: ident: Quantum Radio. Generated commentary on the latest quantum physics and condensed matter papers.

Kai: Today's paper: "Unconditional and exponentially large violation of classicality".

Mira: The gist: The complement sampling game allows for an unconditional and exponentially large violation of classicality when tested on quantum hardware,

Kai: First, who's behind it and why it matters.

Paper summary: Kai: So we're looking at this paper, "Unconditional and exponentially large violation of classicality," which is basically a test for non-classical behavior using something called complement sampling. The main idea here is that this game can show an exponential gap between what a quantum computer can do and what a classical one can manage, without needing any really hard complexity assumptions to make it work one <ref:2511.11008#pg1>.

Mira: Right. It claims this game has all the right ingredients for testing quantum mechanics—it doesn't rely on those unproven complexity-theoretic assumptions, it's easy for a classical computer to verify what happens, and there’s a strategy that is provably exponentially better than the best classical strategy one <ref:2511.11008#pg1>.

Lev: For us running this on real hardware, the paper points out that any strategy succeeding with a probability larger than one-half over many rounds is non-classical one <ref:2511.11008#pg1>. That means we don't need impossibly perfect strategies to prove it's quantum behavior.

Kai: Exactly. They show that for Bernstein-Vazirani subset states, the ideal quantum score is one, while the best classical expected score is much lower, specifically one over two to the power of n minus one fourteen <ref:2511.11008#pg1>. That gives a quantum-classical ratio of two to the power of n minus one fourteen <ref:2511.11008#pg1>.

Mira: It really hinges on this comparison. The paper demonstrates that this gap grows exponentially with the size of the problem, which is what makes it so interesting for testing these kinds of systems. It's not just showing a small difference; it's showing how much power quantum mechanics has in this specific context fifteen <ref:2511.11008#pg1>.

Lev: I wonder about the practical side here. The paper mentions that classical sample complexity to solve complement sampling for BV subsets is about n, but the quantum algorithm gets a success probability of at least one-half plus one over two to the power of n minus q plus one for q samples sixteen <ref:2511.11008#pg1>. That's a bit more concrete on how much work is involved.

Kai: And that's where we bring in the hardware part. The experiment was done on Quantinuum System Model H2 trapped-ion quantum computers, and they tested this with up to fifty-five qubits one. They ran it for n equals thirty-seven and five hundred rounds, and the empirical score stayed significantly above that optimal classical score seven.

Mira: That is a pretty strong result. Reaching significance at alpha equals zero point zero one shows that the device being tested is behaving in a way that's inconsistent with classical physics seven. It’s not just an anomaly; it’s statistically significant evidence for quantum behavior.

Lev: From an error correction standpoint, if you try to run something like this on noisy hardware, you have to account for the noise susceptibility they mention eleven <ref:2511.11008#pg1>. The paper suggests that any strategy succeeding with probability at least one-half plus one over poly n is provably non-classical eight. That gives us a limit on how much noise we can tolerate before the quantum advantage disappears.

Paper summary: Kai: So, while it's resilient to some noise, the core finding is that this game lets us test for non-classicality without relying on those complexity assumptions that usually cloud these kinds of experiments ten <ref:2511.11008#pg1>. It keeps it grounded in verifiable tasks rather than abstract theory.

Mira: I think what makes this specific paper so compelling is how it frames the problem. They’re not talking about entanglement or non-locality, which are the usual suspects in these discussions, but they're testing the power of superposition itself through a specific sampling task two <ref:2511.11008#pg1>. It's a different angle on how we look at quantum mechanics.

Lev: If we think about what this means for real systems, it suggests that even without complex entanglement protocols, you can find measurable evidence of quantum mechanics using simpler, single-player setups two <ref:2511.11008#pg1>. That lowers the barrier for testing these ideas on the hardware we actually have access to.

Kai: So, to summarize this paper "Unconditional and exponentially large violation of classicality," they’ve set up a game where a quantum strategy beats any classical strategy by an exponential margin, and their experiment on fifty-five qubits confirmed that this happens in practice one.

Mira: It really comes down to showing that the advantage isn't just theoretical potential; it's something you can actually measure with current technology under reasonable conditions seven. The implication is that we can use these kinds of games to probe quantum mechanics in ways that are accessible and verifiable.

Lev: Moving forward, the paper flags a limitation regarding how they set up the referee and player roles. They mention needing to make sure both random numbers and those subset states come from their own computer rather than relying on simulated communication channels eight. That kind of setup detail is crucial for making these tests truly robust in a real lab setting.

Kai: That’s the next big thing for testing this idea, I think. Getting rid of any simulated links between the components and having them genuinely independent is what separates a promising result from a fully validated experiment eight.

Mira: It also highlights that while this test is good because it's verifiable, we still need to keep an eye on hardware noise, because that’s always the practical hurdle when you try to prove something non-classical with physical qubits eleven <ref:2511.11008#pg1>.

Lev: Yeah, so the challenge is balancing the theoretical power of this exponential separation against the messy reality of building and maintaining a quantum device eight. It shows what's possible with current hardware, but it also points exactly where the next generation of testing needs to focus.

Conclusion: Kai: So we're wrapping up this look at "Unconditional and exponentially large violation of classicality." It boils down to this game where a quantum approach beats any classical approach by an exponential margin, and they showed it happens on actual hardware.

Mira: Yeah, the authors are using Bernstein-Vazirani states for this test, which is interesting because it's a specific type of quantum state. The core thing here is that they're testing the power of superposition itself, without needing any really complex assumptions about entanglement or non-locality.

Lev: From my side, what I see is that they’ve set up a proof structure where if you can get just a little bit better than a coin flip—one half probability—then the system has to be quantum. That’s the rule for this kind of test.

Kai: Exactly, and the numbers they report are pretty striking. They found that for thirty-seven qubits, running five hundred rounds, their empirical score stayed way above what we'd expect from a classical computer trying its best. It's statistically significant enough to reject the idea that it’s just classical physics on a quantum machine.

Mira: That rejection of the null hypothesis is what really matters here. It means that even on this specific hardware, they’re seeing behavior that simply doesn't fit the classical description of how those states should interact. It points toward a fundamental quantum nature in what the system is doing at a basic level.

Lev: But we gotta be careful with how they set up the referee and player roles. They mentioned needing to make sure both random numbers and those subset states come from their own computer, not some simulated communication channel between them, which is a real practical hurdle for getting this exact result on a lab bench.

Kai: That's the caveat we need to watch. So they've shown the quantum advantage exists in principle, but proving it reliably on real equipment means making sure those inputs are genuinely independent and not secretly linked by some simulation trick.

Mira: And for anyone listening who isn't deep into quantum theory, what this means is that we don't need a whole bunch of complicated entanglement to see evidence of quantum mechanics working in a measurable way. It’s about the state itself being different from anything classical can produce through computation alone.

Lev: So it changes things by showing that even simple sampling games can be powerful tools for probing hardware, provided you handle those input dependencies correctly. We gotta keep looking at these kinds of tests to see where the real practical limitations are hiding in the noise.

Quantinuum

quant-ph

Submitted: 2025-11-14

Updated: 2026-10-08

Comments: Main text: 13 pages, 2 figures, 2 tables. Supplementary information: 15 pages, 5 figures, 3 tables. Published version

Journal ref: Nat. Commun. 17, 10557 (2026)

DOI: 10.1038/s41467-026-77413-3

Code: https://github.com/CQCL/quantinuum-hardware-quantum-volume

License: http://arxiv.org/licenses/nonexclusive-distrib/1.0/

Importance score: 84/100

The gist: The gist: The complement sampling game allows for an unconditional and exponentially large violation of classicality when tested on quantum hardware, demonstrating the power of quantum superposition

Key concepts

Complement Sampling Game
This is a game where you are given a set S and must return a sample from its complement, S̄. The goal is to test if the system can do this better than classical methods by returning an answer that is definitively outside the original set.
Bernstein-Vazirani (BV) Subset States
These are specific types of quantum states used in the experiment. They are prepared based on a random string, and they allow researchers to compare the performance of quantum versus classical strategies for solving sampling problems, providing a concrete test case.
Quantum vs. Classical Ratio
This ratio compares the expected score achieved by an optimal quantum strategy against the expected score of the best possible classical strategy. A high ratio indicates a significant, provable violation of classical physics or computation rules.

Terminology

Summary

The gist: The complement sampling game allows for an unconditional and exponentially large violation of classicality when tested on quantum hardware, demonstrating the power of quantum superposition without relying on computational hardness assumptions.

Complement Sampling Game Formulation

The complement sampling game is a single-player game designed to test non-classicality by returning any sample from the complement set S¯ = ω - S given access to the subset state S⟩ corresponding to a uniform distribution over S (Page 2). The core idea involves swapping the state to its complement S⟩ → S¯⟩ and then measuring it in the computational basis (Page 2). The game is defined by a score function σ(S, y) = (+1, y ∈ S̄, y /∈ S) (Page 3).

Quantum vs. Classical Strategies

The paper compares the performance of classical and quantum strategies against the optimal classical strategy to quantify non-classicality (Page 4). The expected score over the randomness of input and output is defined as V(F, h) = ES∼Unif(F) Ey∼h(yS) [σ (S, y)] (Page 4). For Bernstein-Vazirani (BV) subset states, the ideal quantum value is V(FBV, Q) = 1 because the swapper circuit achieves an expected score of one (Page 13). In contrast, the optimal classical expected score for BV subsets is V(FBV, C) = 1/2(n-1) (Page 14). This leads to a quantum-classical ratio of V(FBV, Q)/V(FBV, C) = 2(n-1) (Page 14).

Implementation and Hardware Experiments

The experiment is executed on Quantinuum System Model H2 trapped-ion quantum computers with up to 55 qubits (Page 1). The game is formulated using BV subsets, where the referee prepares the state S⟩ based on a randomly chosen string u (Page 3). The optimal classical strategy involves measuring the incoming state in the computational basis and returning a different bit string (Page 4). Experimental results show that for n = 37, running for r = 500 rounds, the empirical score Vˆr remains significantly above the optimal classical score V(FBV, C) (Page 7). This rejection of the null hypothesis with significance α = 0.01 corroborates the non-classical nature of the quantum device (Page 7).

Scalability and Noise Resilience

The paper addresses hardware limitations by proposing an alternative family of subsets, BV subsets, which do not rely on unproven complexity-theoretical assumptions (Page 5). The classical sample complexity to solve complement sampling for BV subsets is shown to be Θ(n) (Page 15), while the quantum algorithm achieves a success probability of at least 1/2 + 1/2(n-q+1 - 1) for q samples (Page 16). The experiments are resilient to noise to some extent, as any strategy succeeding with probability at least 1/2 + 1/poly(n) is provably non-classical (Page 8).

Conclusion and Future Directions

The statistics generated exhibit exponentially large violations of classicality, explained by a systematic adoption of a near-optimal quantum strategy (Page 8). The demonstration highlights the power of quantum superposition in a manner oblivious of entanglement and non-locality (Page 8). Future work will focus on resolving loopholes, such as ensuring the referee generates both random numbers and subset states using their own computer rather than relying on simulated communication channels (Page 8). The results provide substantiation of the quantum nature of the hardware used (Page 8).

reference:2511.11008

reference:2511.11008

reference:2511.

Improvements for AI systems

  1. Unconditional Non-Classicality Testing: The improved system can test for quantum mechanical nature by running games like complement sampling, which has an unconditional and exponentially large violation of classicality. This allows for hardware verification that is efficient and scalable way to corroborate the quantum nature of the hardware.

  2. Device-Independent Verification: The system can perform tests where conclusions are drawn solely from the players’ output statistics without any reference to the details of their hardware, making it partially device-independent regarding input preparation, testing the power of quantum superposition rather than just entanglement.

  3. Resilience to Noisy Hardware: By using games like complement sampling based on Bernstein-Vazirani subsets, the system achieves a test that is low hardware requirement and suitable for contemporary noisy hardware, as the ideal quantum strategy succeeds with probability one in an error-free setup.

  4. Scalable Verification Under Noise: The system can handle noise by using statistical testing based on Hoeffding’s inequality to reject the null hypothesis (classical strategy) if the empirical score distance exceeds a threshold, providing a p-value upper bound δ for experimental results.

  5. **High-Fidelity Quantum Simulation/Communication: The system can simulate quantum communication channels between referee and player using quantum teleportation protocols, allowing complex game setups to be implemented on current trapped-ion hardware with overhead that is linear scaling of the number of two-qubit gates.

Sources

Related papers