Experimental Demonstration of the PBR Test on a Superconducting Processor

Basic Information

• Paper ID: 2510.11213
• Title: Experimental Demonstration of the PBR Test on a Superconducting Processor
• Authors: Songqinghao Yang, Haomu Yuan, and Crispin H. W. Barnes (Cambridge University)
• Classification: quant-ph (Quantum Physics)
• Publication Date: October 14, 2025
• Paper Link: https://arxiv.org/abs/2510.11213

Abstract

This paper presents the first experimental verification of the Pusey-Barrett-Rudolph (PBR) no-go theorem on IBM's 156-qubit Heron2 Marrakesh superconducting quantum processor. By preparing non-orthogonal quantum states and evolving them under carefully compiled unitary circuits, the researchers tested whether hidden variable models of quantum states can be interpreted as purely epistemic (reflecting ignorance of some underlying physical reality). To account for real hardware imperfections, the researchers derived noise-aware error tolerances based on decoherence models calibrated to device performance. The results demonstrate that the vast majority of adjacent qubit pairs and adjacent five-qubit configurations produce statistical results that violate epistemic bounds, thereby excluding the epistemic interpretation of quantum mechanics.

Research Background and Motivation

Core Problem

The physical interpretation of quantum states is a central controversy in the foundations of quantum mechanics. Two main perspectives exist:

1. ψ-ontic view: Pure quantum states directly correspond to physical reality
2. ψ-epistemic view: Quantum states represent only knowledge of some underlying physical states

Significance

This question bears on the foundational interpretation of quantum mechanics and affects our understanding of the nature of quantum reality. The PBR theorem provides a crucial no-go result that, under the assumption of preparation independence, proves that any viable hidden variable model cannot be ψ-epistemic.

Existing Limitations

• Previous PBR experiments were primarily conducted in trapped ion and spin systems
• Lack of implementation on scalable, commercially relevant quantum platforms
• Insufficient theoretical modeling of real hardware noise

Research Motivation

1. First implementation of the PBR test on a superconducting qubit architecture
2. Extension of the protocol from 2-qubit to 5-qubit systems
3. Development of a more detailed noise modeling framework
4. Exploration of the PBR test as a quantum character benchmark for NISQ devices

Core Contributions

1. First Superconducting Implementation: First experimental verification of the PBR theorem on IBM's 156-qubit Heron2 processor
2. Systematic Scaling: Extension of the PBR protocol from 2-qubit to 5-qubit configurations
3. Noise-Aware Modeling: Development of a comprehensive noise framework combining depolarization errors and decoherence models based on T₁ and T₂ times
4. Spatial Dependency Analysis: Systematic investigation of the effect of physical distance between qubits on PBR test success rates
5. Device-Level Benchmarking: Demonstration that the PBR test can serve as a quantitative benchmark for quantum character in NISQ systems

Methodology Details

Task Definition

The core of the PBR test is to prepare n systems, each in one of two non-orthogonal states |ψ₀⟩ or |ψ₁⟩, and then perform entangled measurements such that each measurement outcome has zero probability for specific input configurations.

Theoretical Framework

Quantum State Preparation

For single-qubit states:

• |ψ₀⟩ = R_Y(θ)|0⟩ = cos(π/8)|0⟩ + sin(π/8)|1⟩
• |ψ₁⟩ = R_Y(-θ)|0⟩ = cos(π/8)|0⟩ - sin(π/8)|1⟩

where θ = π/4 is the minimum angle satisfying PBR conditions.

Noise Modeling

Depolarization Model:

ρ → (1-p)ρ + p/2 · I

Cumulative error: ε_dep ≈ 1 - (1-p₁)^G₁(1-p₂)^G₂

Thermodynamic Model: Based on T₁ and T₂ times with amplitude damping and pure dephasing:

• p_ad = 1 - exp(-t_g/T₁)
• p_φ = 1 - exp(-t_g/T₂)
• ε_dec ≈ n(p_ad + p_φ)

Error Tolerance

Ideally, the probability of forbidden outcomes should be zero. In a noisy environment, if the probability of each forbidden outcome is bounded by ε_tol, then:

D(μ₀, μ₁) ≥ 1 - 2ε_tol^(1/n)

Circuit Implementation

• For n=2 case, entangling circuits with parameters (α=π, β=0) are used
• For n=5 case, extension to five-qubit configurations
• Qiskit transpiler used to optimize to nearest-neighbor connectivity

Experimental Setup

Hardware Platform

• Device: IBM Heron2 Marrakesh 156-qubit superconducting processor
• Native Gates: CZ, SX, X, RZ, RX, ID, RZZ
• Calibration: Daily acquisition of gate error rates, readout errors, and T₁, T₂ values

Experimental Configuration

1. Adjacent Qubit Pairs (n=2): 20 directly connected qubit pairs
2. Adjacent Five-Qubit Clusters (n=5): 5 different topologies of five-qubit configurations
3. Distant Qubit Pairs: Systematic increase in physical separation between qubits

Measurement Parameters

• 100,000 measurements per circuit execution
• Angle selections: θ = π/4 (minimum) and θ = 1.2π/4 (extended testing)

Experimental Results

Main Results

Adjacent Qubit Pairs (n=2)

• Success Rate: Approximately 97.5% of configurations violate the ψ-epistemic bound
• Error Rates: Best case 0.17±0.02%, worst case 0.56±0.03%
• Theoretical Prediction: Forbidden outcome probability ∼10⁻²-10⁻³ order of magnitude

Five-Qubit Configurations (n=5)

• Successful configurations tend to have higher average T₁, T₂ and fewer two-qubit gates
• Best case: 0.32±0.10%, worst case: 0.48±0.18%

Spatial Dependency Analysis

Significant finding: PBR test pass probability steadily decreases with increasing physical separation between qubits

• Minimum separation: 100% pass rate
• Maximum tested distance (7 intermediate qubits): ~25% pass rate
• Error rate increases from ~0.14% to over 6%

Noise Tolerance Thresholds

Extreme Case Testing

Pairing of qubit 0 with qubit 155 (spanning the entire device) failed the PBR test, with result statistics far exceeding the analyzed tolerance bounds.

Related Work

Historical Development

• PBR Theorem (2012): Foundational no-go theorem proposed by Pusey, Barrett, and Rudolph
• Early Experiments: Primarily conducted in trapped ion and spin systems
• Theoretical Improvements: Multiple works refining experimental and theoretical frameworks

Innovations in This Work

1. First implementation on superconducting qubit architecture
2. Extension to larger system scales
3. Comprehensive noise modeling accounting for real hardware
4. Systematic investigation of spatial connectivity effects

Conclusions and Discussion

Main Conclusions

1. Successful Verification: PBR test successfully implemented on superconducting platform, excluding the ψ-epistemic interpretation
2. Noise Sensitivity: Protocol is highly sensitive to circuit depth, gate count, and qubit separation
3. Benchmarking Potential: PBR test can serve as a rigorous benchmark for coherence and gate fidelity in NISQ devices

Physical Interpretation

From an epistemic perspective, decoherence is an information process leading to gradual overlap of hidden variable distributions μ₀(λ) and μ₁(λ). From an ontic perspective, the wavefunction is a genuine physical entity, and unitary evolution redistributes coherence without inducing overlap.

Limitations

1. Incomplete Noise Models: Simple noise models become inadequate for large-scale systems
2. Circuit Depth Constraints: Cumulative errors in deeper circuits exceed current model predictions
3. Theoretical Controversies: Recent theoretical critiques point to potential limitations in the PBR framework

Future Directions

1. Extension to larger n or different state angles
2. Application to benchmarking new devices
3. More precise characterization of correlated and non-Markovian noise processes
4. Comparative studies with quantum volume and cross-entropy benchmarks

In-Depth Evaluation

Strengths

1. Pioneering Work: First implementation of PBR test on commercially relevant superconducting platform
2. Comprehensive Noise Analysis: Multiple noise models providing realistic error bounds
3. Systematic Investigation: Complete testing from adjacent to distant qubits
4. Practical Value: New benchmarking methodology for NISQ devices

Weaknesses

1. Theoretical Model Limitations: Limited predictive power of noise models for large-scale systems
2. Declining Success Rates: Significant decrease in success rates with increasing system size and distance
3. Device Dependence: Results highly dependent on specific hardware performance characteristics

Impact

1. Fundamental Physics: Provides experimental evidence for quantum mechanics interpretation debates
2. Quantum Computing: New evaluation tool for NISQ devices
3. Technology Development: Drives advancement in noise modeling and error correction techniques

Applicable Scenarios

1. Device-level benchmarking of quantum processors
2. Assessment of quantum coherence and gate fidelity
3. Experimental platform for fundamental quantum mechanics research
4. Reference standard for NISQ algorithm performance prediction

References

This paper cites 21 relevant references covering the theoretical foundations of the PBR theorem, early experimental implementations, noise modeling theory, and important works in quantum computing benchmarking methods.

This work represents not only a significant technical breakthrough but also provides a new experimental perspective on fundamental questions regarding the interpretation of quantum mechanics, while opening new directions for the evaluation of quantum devices in the NISQ era.