Quantum computers reveal that the wave function is a real thing

Does quantum mechanics really reflect nature in its truest form, or is it just our imprecise way of describing the weird properties of the very small? A famous test that can help answer this question has now been tried on a quantum computer, and it comes to a surprisingly concrete conclusion. Quantum mechanics really does describe reality completely, at least for tiny quantum devices – and the results could help us build better and more reliable quantum machines.

Since quantum mechanics was first discovered more than a century ago, its uncertain and probabilistic nature has troubled physicists. Take, for example, a superposition – is a particle actually inhabiting many places at once, or is the calculation of its position giving us a range of probabilities for where it actually is? If it is the latter, there may be some feature of reality that is hidden to quantum mechanics that limits our certainty. Such a feature would be a “hidden variable”, and so theories predicated on this idea are called hidden variable theories.

In the 1960s, physicist John Bell devised an experiment to rule out such theories. A Bell test probes quantumness by measuring how linked, or entangled, distant pairs of quantum particles are. If their quantum properties are maintained above a certain threshold – if their entanglement is what we call non-local, spanning any distance – then we could rule out hidden variable theories. Bell tests have since been tried for many quantum systems, unanimously ruling in favour of the inherent non-locality of the quantum world.

In 2012, physicists Matthew Pusey, Jonathan Barrett and Terry Rudolph came up with an even more probing test (named PBR after them), which would allow experimenters to differentiate between various interpretations of a quantum system. These include the ontic view, which says our measurements of a quantum system and its wave function – the mathematical description of its quantum states – represent reality. Another interpretation, called the epistemic view, says this wave function is a mirage and there exists a deeper, richer reality underneath.

Assuming you believe that quantum systems don’t have some other secret feature that can affect systems beyond the wave function, then the mathematics of the PBR show that you should always get an ontic view of things – that however weird they may look, quantum behaviours are real. The PBR test works by comparing different quantum elements, such as a qubit inside a quantum computer, and measuring how often they read out the same value for some property, such as their spin. If the epistemic view were correct, the amount of times that your qubits read the same value would be higher than quantum mechanics predicts, indicating something else is going on underneath.

Songqinghao Yang at the University of Cambridge and his colleagues have devised a way to carry out the PBR test on a working IBM Heron quantum computer, and they saw that for small numbers of qubits, we can indeed say that quantum systems are ontic. That is, quantum mechanics appears to work as we thought, just as Bell tests have repeatedly found.

Yang and his team carried out this check by measuring the overall output produced by pairs or groups of five qubits, such as strings of 1s and 0s, and calculated how often this result lined up with their prediction of how a quantum system should behave, accounting for the natural errors in the system.

“Currently, all quantum hardware is noisy, and there are some errors on all operations, so if we add in this noise on top of the PBR threshold, then what would happen to our interpretation [of our system]?” says Yang. “It turns out that if you do the experiment on a small scale, then we can still satisfy the original PBR test and we can rule out the epistemic interpretation.” Hidden variables, be gone.

While they could show this for small numbers of qubits, they struggled to do the same for larger numbers of qubits on the 156-qubit IBM machine. The noise, or errors, in the system became too great for the researchers to distinguish between the two scenarios in a PBR test.

This means the test can’t tell us if the world is quantum all the way up. It could be that at some scales, the ontic view wins out, while at larger scales we aren’t able to see precisely what quantum effects are doing.

Being able to verify a quantum computer’s “quantumness” using this test could be a way to confirm that these devices are doing what we think they are, as well as make them more likely to be able to display a quantum advantage – the ability to do a task that would take a classical computer an unreasonable amount of time. “If you want to have quantum advantage, you need to have quantumness inside your quantum computers, or else you can find an equivalent classic algorithm,” says team member Haomu Yuan at the University of Cambridge.

“The idea of using PBR as a benchmark of device performance is intriguing,” says Matthew Pusey at the University of York, UK, one of PBR’s original authors. But Pusey is less sure that it is telling us something about reality. “The main reason to do the experiment, rather than relying on theory, is if you think quantum theory could be wrong. But if quantum theory is wrong, what question are you even asking? The whole setup of ontic vs epistemic states presupposes quantum theory.”

To truly find a way to do a PBR test that would tell us about reality, you would need to find a way to do it without presupposing quantum theory is correct. “There are a minority of people who believe that quantum physics will fundamentally break down at some mesoscopic scale,” says Terry Rudolph at Imperial College London, another of the originators of the PBR test. “While this experiment is not likely relevant to ruling out any specific such proposal out there – to be clear, I don’t know one way or the other! – testing the fundamental features of quantum theory on ever larger systems always helps us narrow the search space of alternative theories.”

Reference: arXiv, DOI: arxiv.org/abs/2510.11213