IBM and Rutgers Control Quantum Chaos Across 100 Qubits on Heron Processor

Controlling Quantum Chaos on a 100‑Qubit Processor

Researchers from Rutgers University, IBM Quantum, and partner institutions have shown that adaptive quantum circuits can tame chaotic dynamics across a large superconducting processor. By interleaving entangling gates with mid‑circuit measurements and resets on 100 linearly connected qubits of IBM’s 156‑qubit Heron chip, they observed a controllable phase transition between a scrambling, volume‑law entangled regime and a reset‑dominated, unentangled regime. The work, published in Nature Physics, provides the first experimental validation of non‑unitary dynamics that are essential for scalable quantum error correction.

---

1. Adaptive Circuits and the Control Parameter

The experiment implements a repeated sequence of two‑qubit entangling gates followed by mid‑circuit measurements and conditional resets. A single control knob, the reset probability p, determines how often a qubit is forced back to the ground state after measurement. Varying p tunes the competition between:

• Scrambling unitaries that spread quantum information (low p), and
• Local resets that erase information and drive the system toward a fixed state (high p).

Four dynamical regimes emerge:

| Regime | Reset range | Entanglement behavior | Lyapunov exponent | |--------|-------------|----------------------|-------------------| | Chaotic Volume‑Law | 0 ≤ p < 0.3 | Extensive volume‑law entanglement | λ > 0 | | Chaotic Area‑Law | 0.3 < p < 0.5 | Sub‑extensive, area‑law scaling | λ ≈ 0 | | Critical Point (CIPT) | p ≈ 0.5 | Balanced scrambling and resets; diffusive magnetization growth | λ ≈ 0 | | Controlled Phase | 0.5 < p ≤ 1 | Unentangled, system locked to |0…0⟩ | λ < 0 |

At the critical reset probability pₙ ≈ 0.495, the team measured critical exponents (ν ≈ 1.05, z ≈ 2.08, β ≈ 1.05) consistent with theoretical predictions for a measurement‑induced phase transition.

---

2. Experimental Scale and Fidelity

• Qubits used: 100 linearly connected qubits on the Heron processor.
• Gate count: ~5,000 two‑qubit entangling gates across the full circuit depth.
• Measurements & resets: ~5,000 mid‑circuit measurement‑reset operations.
• Shots per configuration: 10,000 repetitions to gather reliable statistics.

The hardware achieved average two‑qubit gate fidelities above 99 % and measurement fidelities near 98 %, enabling the delicate balance required to observe the transition. The large qubit count pushes the limits of current superconducting platforms, demonstrating that real‑time feedback can be scaled beyond the few‑qubit demonstrations of earlier work.

---

3. Verifying Genuine Quantum Dynamics

To ensure the observed behavior stems from coherent quantum processes rather than classical noise, the researchers analyzed the shot‑to‑shot variance of the magnetization observable \(M_z\). By comparing the experimental distribution to a classical dephasing Markov model using Kullback‑Leibler divergence, they confirmed a statistically significant quantum signature. This quantum‑fluctuation sampling also serves as a hardware certification tool, showing that the processor can faithfully implement non‑unitary, measurement‑driven dynamics.

---

4. Implications for Quantum Error Correction

Measurement‑induced phase transitions are a cornerstone of many proposed error‑correction schemes, where frequent syndrome measurements collapse errors while preserving logical information. Demonstrating control over the transition on a 100‑qubit substrate validates key assumptions of these protocols:

• Scalable feedback: Mid‑circuit measurements and conditional resets can be performed reliably at depth scales required for surface‑code cycles.
• Non‑unitary dynamics: Real hardware can implement the stochastic processes modeled by matrix‑product‑state simulations and statistical‑mechanics mappings.
• Phase‑boundary tuning: Adjusting measurement rates offers a practical knob for optimizing error‑suppression versus overhead.

While the study does not yet implement a full logical qubit, it establishes a concrete experimental platform for testing next‑generation error‑correction architectures.

---

5. Looking Ahead

The success of this 100‑qubit feedback experiment opens several avenues for future work:

1. Extending to 2‑D connectivity: Moving beyond linear chains to planar lattices will more closely match surface‑code layouts.
2. Integrating logical operations: Embedding logical gates within the adaptive circuit will test fault‑tolerance under realistic workloads.
3. Hybrid architectures: Combining superconducting qubits with fast classical processors for low‑latency control could further reduce measurement overhead.

As quantum hardware continues to scale, the ability to steer chaotic dynamics into ordered, error‑correctable regimes will be a decisive factor in achieving practical quantum advantage.

---

This article synthesizes the findings reported in the recent Nature Physics paper and related institutional releases, focusing on the experimental methodology, observed phase transition, and broader relevance to quantum computing.