Wigner's Friend as a Circuit: Inter-Branch Communication Witness Benchmarks on Superconducting Quantum Hardware

Abstract: We implement and benchmark on IBM Quantum hardware the circuit family proposed by Violaris for estimating operational inter-branch communication witnesses, defined as correlations in classical measurement records produced by compiled Wigner's-friend-style circuits. We realize a five-qubit instance of the protocol as an inter-register message-transfer pattern within a single circuit, rather than physical signaling, and evaluate its behavior under realistic device noise and compilation constraints. The circuit encodes branch-conditioned evolution of an observer subsystem whose dynamics depend on a control qubit, followed by a controlled transfer operation that probes correlations between conditional measurement contexts. Executing on the ibm_fez backend with 20000 shots, we observe population-based visibility of 0.877, coherence witnesses of 0.840 and -0.811 along orthogonal axes, and a phase-sensitive magnitude of approximately 1.17. While the visibility metric is insensitive to some classes of dephasing, the coherence witnesses provide complementary sensitivity to off-diagonal noise. This work does not test or discriminate among interpretations of quantum mechanics. Instead, it provides a reproducible operational constraint pipeline for evaluating detectability of non-ideal channels relative to calibrated device noise.

• The paper presents a hardware benchmark using a five-qubit branch-transfer circuit to evaluate coherence witnesses and inter-branch communication on IBM's superconducting quantum backend.
• It employs both population visibility and multi-qubit Pauli correlators (W_X, W_Y) to quantify circuit performance, revealing a phase magnitude of 1.1673 versus an ideal 1.414.
• The study establishes a reproducible framework to constrain nonunitary deviations, paving the way for cross-platform quantum foundational tests and refined error-mitigation strategies.

Benchmarking Wigner's Friend Circuit Protocols: Coherence Witnesses and Inter-Branch Communication on Superconducting Quantum Hardware

Introduction

This paper presents the first hardware benchmark of circuit families inspired by Violaris (2026), explicitly encoding Wigner's-friend-style scenarios to evaluate operational "inter-branch communication" witnesses using superconducting qubit devices. The experiment transposes the abstract Wigner's friend scenario—traditionally a conceptual probe of quantum observation and measurement—into a parameterized five-qubit quantum circuit, rigorously executed on IBM Quantum's "ibm_fez" backend. Focus is placed on coherence-sensitive correlators beyond population-based diagnostics to assess how branch-conditioning and circuit compilation manifest in realistic device environments.

Circuit Protocol and Measurement Methodology

The branch-transfer primitive employs five logical qubits, mapping four of these ($Q$, $R$, $F$, $P$) to measured roles—with $Q$ serving as branch-control, $F$ as "friend," $R$ as a reference for message encoding, and $P$ as a probe. The protocol comprises three stages: branch-conditioned unitary evolution of $F$, a controlled message-transfer operation mediating register correlations, and selective measurement in either the computational basis ($V$ visibility) or rotated bases for coherence witnesses ($W_X$, $W_Y$).

The schematic as transpiled for hardware execution:

Figure 1: Branch-transfer circuit schematic (five-qubit primitive) highlighting measurement and branch-transfer roles following device transpilation.

Population visibility ($V$) is extracted from $R$-$P$ diagonal measurement statistics. Its main limitation—vulnerability to dephasing channels that destroy off-diagonal coherence without perturbing diagonal populations—necessitates additional benchmarking via multi-qubit Pauli correlators. The $W_X$ and $W_Y$ witnesses quantify the expectation values of joint $X$- and $Y$-parity operators across $Q$, $R$, $F$, $P$, providing operational sensitivity to coherent branch interference. The phase-sensitive magnitude $C_{\mathrm{mag}} = \sqrt{W_X^2 + W_Y^2}$ provides a rotationally invariant coherence estimate; for the ideal circuit, $C_{\mathrm{mag}} = \sqrt{2}$.

Hardware Execution and Backend-Matched Simulation

Experiments use 20,000 shots per circuit, with hardware runs at transpiler optimization level 2 and simulations at level 1 for alignment with calibration constraints. A provenance bundle archives job IDs, calibration snapshots, and software versions, ensuring reproducibility. Noise modeling is conducted using Qiskit Aer, leveraging contemporaneous calibration data for error rates, coherence times, and readout infidelities.

Simulated and hardware benchmarks are compared across all metrics, utilizing backend-derived models as proxies but not strict predictions. Numerical results derived from both hardware and simulation establish the baseline noise floor and observable degradation.

Results: Empirical Witness Benchmarks

The primary hardware measurement delivers:

• Population visibility $V = 0.8771 \pm 0.0034$
• Coherence witnesses $W_X = 0.8398 \pm 0.0038$, $W_Y = -0.8107 \pm 0.0041$
• Phase magnitude $C_{\mathrm{mag}} = 1.1673 \pm 0.0040$

These values exhibit statistically significant degradation from the ideal ($V = 1.0000$, $C_{\mathrm{mag}} = 1.414$), with hardware performance below backend-matched noisy simulation, implicating calibration drift, gate errors, or compilation overheads beyond those modeled.

Figure 2: Coherence witnesses ($W_X$, $W_Y$, $C_{\mathrm{mag}}$) across ideal, backend-matched simulation, and hardware runs.

Comparisons across measurement modalities reveal the necessity of coherence witnesses—hardware visibility remains relatively insensitive to certain nonunitary channels, while $W_X$ and $W_Y$ provide direct constraints on off-diagonal coherence, enabling channel detectability analysis.

Figure 3: Ideal baseline for the branch-transfer circuit, illustrating maximal coherence and visibility.

Nonunitary Channel Constraint Pipeline

Observables are used to define exclusion bounds for parameterized nonunitary channel families (e.g., single-qubit dephasing). The constraint protocol:

1. Inserts a channel (dephasing parameter $\lambda$) at a specified circuit location.
2. Simulates the induced effect on $V$, $W_X$, $W_Y$, and $C_{\mathrm{mag}}$ vs. $\lambda$.
3. Defines detectable $\lambda$ where predicted deviations exceed combined uncertainty and noise bands.
4. Empirically constrains $\lambda$ on the given hardware.

An explicit channel example ($\mathcal{E}_\lambda$: phase-flip on $F$) reveals coherence witness attenuation with increasing $\lambda$, yielding illustrative hardware-constrained bounds $\lambda_\mathrm{est} \approx 0.08$ for measured $W_X$.

Backend-matched forecasts, both for the ideal and for calibrated noisy baselines, indicate that coherence witnesses provide sharper exclusion limits than visibility alone.

Theoretical and Practical Implications

The experiment demonstrates that branch-transfer primitives and their coherence-sensitive diagnostics exhibit performance consistent with unitary quantum mechanics plus device noise. No anomalous deviations are detected; distinctions between interpretations (Many-Worlds, Copenhagen, RQM, etc.) remain operationally indistinguishable for these circuits when evolution is unitary. However, the developed methodology robustly positions coherence witnesses for future constraints on nonunitary models and collapse-like perturbations.

From an implementation perspective, the work establishes an extensible benchmarking protocol—fully provenance-tracked and reproducible—that can be executed across hardware platforms and compilation strategies.

Prospects for Future Research

Immediate extensions include cross-platform replications (ion traps, neutral atoms, photonic clusters), systematic calibration cycles, refined error-mitigation pipelines, and scaling benchmarks tracking the effect of increased branch divergence (multi-qubit Hamming separation). Quantifying depth overhead, fidelity loss, and error-correction tax as a function of hardware connectivity and compilation complexity will inform both practical deployment of quantum foundational tests and the development of generalized scaling laws for inter-branch communication circuits.

Coherence witnesses remain the primary tool for detecting small nonunitary deviations, especially as device performance and calibration improve.

Conclusion

This study executes and benchmarks a Wigner's-friend-inspired branch-transfer circuit on superconducting qubit hardware, confirming quantitative coherence witness values consistent with unitary predictions within device noise. The developed protocol provides a reproducible, extensible platform for evaluating future candidate nonunitary mechanisms and hardware-specific deviations, establishing operational pipelines for cross-modality investigation in quantum circuit foundations.