We characterize the geometric and algebraic conditions governing noise-induced landscape regularization in Recurrent Quantum Circuits (Q-RNNs). In quantum recurrent models training on sequence parity (temporal XOR), gradient optimization typically encounters a severe topological trap (the parity plateau at accuracy ≈ 0.72–0.75). While physical environmental noise is conventionally viewed as detrimental to quantum information, we provide empirical evidence that the primary boundary separating regularizing and non-regularizing open-system channels is unitality. Specifically: 1. Unital Channels (e.g., Pauli depolarizing, phase-flip, bit-flip, and phase damping) contract the Bloch sphere symmetrically, preserving the loss landscape's up/down parity symmetry and enabling optimizers to escape the topological traps. 2. Non-Unital Channels (e.g., amplitude damping) shift the center of the Bloch sphere; large non-unital origin drift severely disrupts the optimization landscape by breaking the loss landscape's up/down parity symmetry. 3. We map this boundary empirically by conducting a Generalized Amplitude Damping (GAD) sweep over the thermal parameter [0.0, 1.0] across varying damping rates \0.25, 0.35, 0.50, 0.75\ (Sweep 1), and a state-initialization asymmetry sweep (Sweep 2). We show that convergence is restored within a near-unital basin (e.g., [0.42, 0.56] at =0.50). We show that the slight asymmetry of this basin (which is invariant to the initial state of the memory qubit) is consistent with a structural loop asymmetry caused by the recurrent input qubit reset to |0. We also map the parametric scaling of this basin, showing that the critical displacement boundary |v_z, crit| shrinks as increases due to the rapid contraction of the available state space volume. 4. We characterize the fine-grained transition of the bit-flip channel, revealing a sharp degeneracy hole at exactly p=0.50 where the Bloch sphere contracts to a 1D line. 5. We characterize an empirical optimizer-class robustness pattern at this finite experimental depth (2-qubit, T=3): normalized direction-preserving optimizers (Adam, signSGD, Muon, Normalized GD) achieve 6/6 convergence under unital noise, whereas non-normalized gradient descent (SGD, Heavy Ball) fails (2/6) due to recurrent gradient magnitude contraction. We discuss the analogy between this regularization structure and Quantum Error Correction (QEC), noting that the symmetries protecting parity in stabilizer codes appear to play a parallel role in protecting parity classification landscapes under unital noise. All results are from 2-qubit Q-RNN simulations on T=3 parity; generalization to larger circuits and longer sequences is a target for future work.
We characterize the geometric and algebraic conditions governing noise-induced landscape regularization in Recurrent Quantum Circuits (Q-RNNs). In quantum recurrent models training on sequence parity (temporal XOR), gradient optimization typically encounters a severe topological trap (the parity plateau at accuracy ≈ 0.72–0.75). While physical environmental noise is conventionally viewed as detrimental to quantum information, we provide empirical evidence that the primary boundary separating regularizing and non-regularizing open-system channels is unitality. Specifically: 1. Unital Channels (e.g., Pauli depolarizing, phase-flip, bit-flip, and phase damping) contract the Bloch sphere symmetrically, preserving the loss landscape's up/down parity symmetry and enabling optimizers to escape the topological traps. 2. Non-Unital Channels (e.g., amplitude damping) shift the center of the Bloch sphere; large non-unital origin drift severely disrupts the optimization landscape by breaking the loss landscape's up/down parity symmetry. 3. We map this boundary empirically by conducting a Generalized Amplitude Damping (GAD) sweep over the thermal parameter [0.0, 1.0] across varying damping rates \0.25, 0.35, 0.50, 0.75\ (Sweep 1), and a state-initialization asymmetry sweep (Sweep 2). We show that convergence is restored within a near-unital basin (e.g., [0.42, 0.56] at =0.50). We show that the slight asymmetry of this basin (which is invariant to the initial state of the memory qubit) is consistent with a structural loop asymmetry caused by the recurrent input qubit reset to |0. We also map the parametric scaling of this basin, showing that the critical displacement boundary |v_z, crit| shrinks as increases due to the rapid contraction of the available state space volume. 4. We characterize the fine-grained transition of the bit-flip channel, revealing a sharp degeneracy hole at exactly p=0.50 where the Bloch sphere contracts to a 1D line. 5. We characterize an empirical optimizer-class robustness pattern at this finite experimental depth (2-qubit, T=3): normalized direction-preserving optimizers (Adam, signSGD, Muon, Normalized GD) achieve 6/6 convergence under unital noise, whereas non-normalized gradient descent (SGD, Heavy Ball) fails (2/6) due to recurrent gradient magnitude contraction. We discuss the analogy between this regularization structure and Quantum Error Correction (QEC), noting that the symmetries protecting parity in stabilizer codes appear to play a parallel role in protecting parity classification landscapes under unital noise. All results are from 2-qubit Q-RNN simulations on T=3 parity; generalization to larger circuits and longer sequences is a target for future work.
Very little is known about the cost landscape for parametrized Quantum Circuits (PQCs). Nevertheless, PQCs are employed in Quantum Neural Networks and Variational Quantum Algorithms, which may allow for near-term quantum advantage. Such applications require good optimizers to train PQCs. Recent works have focused on quantum-aware optimizers specifically tailored for PQCs. However, ignorance of the cost landscape could hinder progress towards such optimizers. In this work, we analytically prove two results for PQCs: (1) We find an exponentially large symmetry in PQCs, yielding an exponentially large degeneracy of the minima in the cost landscape. Alternatively, this can be cast as an exponential reduction in the volume of relevant hyperparameter space. (2) We study the resilience of the symmetries under noise, and show that while it is conserved under unital noise, non-unital channels can break these symmetries and lift the degeneracy of minima, leading to multiple new local minima. Based on these results, we introduce an optimization method called Symmetry-based Minima Hopping (SYMH), which exploits the underlying symmetries in PQCs. Our numerical simulations show that SYMH improves the overall optimizer performance in the presence of non-unital noise at a level comparable to current hardware. Overall, this work derives large-scale circuit symmetries from local gate transformations, and uses them to construct a noise-aware optimization method.
We revisit an established Quantum Control objective, characterize its exact local metric geometry, and reposition it as a rigorous benchmark for Search and Optimization. The trap-free topology of the landscape, established two decades ago, guarantees unhindered optimal pathways, but global topology alone does not govern convergence speed. Although the so-called Quantum Sphere is defined by the fourth power of the control field, we show that its yield gap admits an exact representation as a nonnegative linear combination of sin2 terms. From this single identity, we derive a two-sided quadratic sandwich with ratio 4/π2 on a certified, worst-case quarter-wave region, proving the optimal peak nondegenerate modulo its affine gauge and well-conditioned. For white-box optimization, on a halved (eighth-wave) region, the objective satisfies a Polyak-Łojasiewicz inequality with an explicit constant, granting Gradient Descent a linear rate of 1−16/(π4κr), where κr is the condition number of the relevant Hessian block. For black-box optimization, we prove that the elitist single-parent Evolution Strategy converges linearly, almost surely and in expected hitting time, from a certified sublevel set, despite the landscape's continuous gauge symmetry and its lattice of periodic optima; the certified level is explicit whenever the landscape is trap-free, and near the optimum the strategy provably sees only the quadratic tangent of the peak. Systematic numerical simulations corroborate the theory. Finally, we show that additive input (actuator) noise reweights each term of the representation by a Debye-Waller factor without disrupting the landscape's structure, and outline open challenges: output (signal) noise, far-field plateau escape, angular-synchronization certificates, and another physically scalable benchmark (SHG with Glass).
Variational quantum algorithms (VQAs) hold much promise but face the challenge of exponentially small gradients. Unmitigated, this barren plateau (BP) phenomenon leads to an exponential training overhead for VQAs. Perhaps the most pernicious are noise-induced barren plateaus (NIBPs), a type of unavoidable BP arising from open system effects, which have so far been shown to exist for unital noise maps. Here, we generalize the study of NIBPs to more general completely positive, trace-preserving maps, investigating the existence of NIBPs in the unital case and a class of non-unital maps we call Hilbert-Schmidt (HS)-contractive. The latter includes amplitude damping. We identify the associated phenomenon of noise-induced limit sets (NILS) of the VQA cost function and prove its existence for both unital and HS-contractive non-unital noise maps. Along the way, we extend the parameter shift rule of VQAs to the noisy setting. We provide rigorous bounds in terms of the relevant variables that give rise to NIBPs and NILSs, along with numerical simulations of the depolarizing and amplitude-damping maps that illustrate our analytical results.
Noisy Intermediate-Scale Quantum (NISQ) devices impose structural constraints on the training of Variational Quantum Algorithms (VQAs). In the absence of full error correction, each quantum gate introduces a non-zero probability of error that accumulates with circuit depth, while gradient estimation through finite sampling adds additional statistical variability. As a result, convergence depends on a delicate balance between physical coherence and estimation variance. In this setting, the depth of the variational ansatz not only determines model expressivity, but also its operational feasibility under noise. Without an explicit characterization of the interaction between error accumulation, number of measurements (shots), and control strategies, increasing experimental resources may fail to improve performance and can even become counterproductive. In this work, we empirically analyze how circuit depth affects the practical signal of the gradient under depolarizing noise. Across multiple configurations (q=4−7qubits), we observe a pattern consistent with an effective exponential attenuation of the coherent gradient contribution, characterized by a per-layer rateϵeff. The estimated magnitude ofϵeffdepends on circuit complexity and remains largely independent of the measurement budget, indicating that it primarily reflects architectural and physical factors rather than sampling effects. The analysis of adaptive control strategies further suggests that accuracy gains are constrained by gradient attenuation, rather than scaling significantly with circuit depth. However, a crossmode comparison across three distinct training strategies reveals that, while accuracy improvements remain modest, adaptive perturbations substantially improve the statistical detectability of the attenuation signal, acting as diagnostic probes rather than mere optimization heuristics. Taken together, these results suggest that the usable depth of a variational ansatz can be interpreted as an emergent property of the effective signal-to-noise regime, governed by a measurable structural parameter whose ranking across circuit configurations is preserved independently of the training mode. This framework enables principled comparison of circuit architectures in terms of their effective trainability under noise, providing guidance for architectural and control design in Quantum Machine Learning under NISQ conditions.
Noisy Intermediate-Scale Quantum (NISQ) devices show great promise for quantum advantage via Parametrized Quantum Circuits (PQCs) and Variational Quantum Algorithms (VQAs). Due to hardware imperfections, however, especially barren plateaus and environmental Pauli noise, the cost landscape is highly distorted, resulting in inaccurate classical optimization during training. To overcome these challenges, a robust quantum gradient descent framework is presented with the integration of error mitigation directly within the optimization loop. A noise-resilient parameter-shift rule is defined and modeled analytically, as is a variant of stochastic gradient descent with real-time Pauli noise mitigation. The framework is based on a 6-qubit system, which is tested on the Max-Cut problem in the presence of different amounts of depolarizing noise. It is shown experimentally that the proposed algorithm is able to achieve the final cost error of 0.045 under significant noise, whereas the standard quantum gradient descent fails to achieve better than 0.320 under this noise. Moreover, the methodology is consistent, converging without a noise-induced flat landscape and without local minima. The method offers a scalable route towards future applications of quantum machine learning on non-fault-tolerant hardware that does not suffer from the large sampling overhead often associated with conventional error mitigation techniques.
Standard Zero-Noise Extrapolation (ZNE) in Variational Quantum Eigensolvers (VQE) assumes a single global noise scaling parameter, failing to capture heterogeneous and non-linear noise propagation across layered quantum circuits. We simulate VQE parameter optimisation at a challenging L=6 hardware-efficient depth, extracting the exact expectation values at each iteration. When the optimal parameters theta_star are found, they are fixed, and the multi-parameter noise surface is mapped. We propose a multi-parameter noise framework, modeling expectation values as a function of layer-wise noise scaling factors E(lambda_1, lambda_2, ..., lambda_L). Using structured unitary folding U_l = U_l (U_l^dagger U_l)^((lambda_l - 1) / 2), we independently scale the noise contribution of each ansatz layer and empirically map the resulting multi-dimensional noise surface. We analyze the separability of inter-layer noise interactions, identify layers with disproportionate error contributions via the gradient dE/dlambda_l, and compare scalar ZNE with multivariate polynomial regression and Gaussian Process (GP) extrapolation across H_2, LiH, and BeH_2 under depolarizing, amplitude damping, and coherent over-rotation noise. Our results demonstrate that noise is non-uniform and partially coupled across layers, and that layer-resolved ZNE with higher-order extrapolation can improve accuracy and variance under realistic NISQ noise models. These findings provide geometric insight into noise propagation in variational circuits and motivate targeted, layer-specific error mitigation strategies.