Abstract We introduce and test two measurement-free training primitives for quantum reservoir computers (QRC). Protocol~A encodes the loss gradient in an auxiliary phase that is fed back coherently before a dissipative reset; Protocol~B realizes gradient descent as relaxation of an engineered open system. Both match exact-gradient training in simulation---the ancilla and reset simulated as channels in a de-risk stage, their validated error model applied at the benchmark stage---on a minimal transverse-field Ising reservoir, with the coherent cost oracle emulated classically and its realization costed. A fully coherent update loop (quantum parameter register, unitary controlled update, no readout anywhere) descends the true landscape, its feedback drift attenuated by exactly $2F-1$ with the fixed point unchanged. Benchmarked against classical reservoir computing--- including the memory-augmented hybrid---they improve one-step chaotic prediction, though a thirty-seed study finds no autonomous-prediction advantage; we therefore make no quantum-advantage claim. The study instead addresses three questions that gate this route. First, within the channel model studied, the reset fidelity already demonstrated in measurement-free quantum error correction [Butt \textit{et al.}, Nat.\Commun.\\textbf{17}, 995 (2026)] suffices: reset is not the bottleneck. Second, the binding constraint is coherent cost-oracle precision, set by a finite-difference catastrophic cancellation that the exact parameter-shift rule removes, loosening the oracle-precision target $2.5$-fold in a like-for-like comparison. Third, a fault-tolerant realization---arithmetic and synthesis costs taken from real circuits---needs only $\sim\!10^{4}$ physical qubits and is runtime-, not qubit-bound. On three free IBM Quantum $156$-qubit Heron backends at three error-mitigation levels the simulation-optimized state survives device noise (RMSE $0.016$--$0.057$, i.e.\$7$--$26\%$ of the signal spread, correlation $\ge0.97$). All code, seeds, and job identifiers are released.
Abstract We introduce and test two measurement-free training primitives for quantum reservoir computers (QRC). ProtocolA encodes the loss gradient in an auxiliary phase that is fed back coherently before a dissipative reset; ProtocolB realizes gradient descent as relaxation of an engineered open system. Both match exact-gradient training in simulation---the ancilla and reset simulated as channels in a de-risk stage, their validated error model applied at the benchmark stage---on a minimal transverse-field Ising reservoir, with the coherent cost oracle emulated classically and its realization costed. A fully coherent update loop (quantum parameter register, unitary controlled update, no readout anywhere) descends the true landscape, its feedback drift attenuated by exactly 2F−1 with the fixed point unchanged. Benchmarked against classical reservoir computing--- including the memory-augmented hybrid---they improve one-step chaotic prediction, though a thirty-seed study finds no autonomous-prediction advantage; we therefore make no quantum-advantage claim. The study instead addresses three questions that gate this route. First, within the channel model studied, the reset fidelity already demonstrated in measurement-free quantum error correction [Butt \textit{et al.}, Nat.\Commun.\textbf{17}, 995 (2026)] suffices: reset is not the bottleneck. Second, the binding constraint is coherent cost-oracle precision, set by a finite-difference catastrophic cancellation that the exact parameter-shift rule removes, loosening the oracle-precision target 2.5-fold in a like-for-like comparison. Third, a fault-tolerant realization---arithmetic and synthesis costs taken from real circuits---needs only ∼104 physical qubits and is runtime-, not qubit-bound. On three free IBM Quantum 156-qubit Heron backends at three error-mitigation levels the simulation-optimized state survives device noise (RMSE 0.016--0.057, i.e.$7−−26%ofthesignalspread,correlation\ge0.97$). All code, seeds, and job identifiers are released.
Quantum reservoir computing (QRC) uses the dynamics of a fixed or weakly tuned quantum system to transform temporal and sequential inputs into measured features, while training is typically confined to a classical readout. This separation reduces reliance on repeated quantum parameter updates and avoids the barren plateaus associated with variational circuit training. Its computational power is often attributed to the exponentially large Hilbert space of the quantum system. However, the memory, nonlinearity, and expressivity that determine what a reservoir can actually compute depend jointly on the input encoding, quantum evolution, observables, measurement, and readout, not on Hilbert space dimension alone. On hardware, these capabilities are further constrained by finite sampling, hardware noise, measurement backaction, and the cost of estimating observables, so a large state space alone does not guarantee useful computation. In this survey, we develop a common system model that connects these components and use it to organize QRC foundations, computational properties, reservoir architectures, operating protocols, and physical implementations. We examine spin, photonic, superconducting, bosonic, neutral atom, and other analog platforms, together with applications, software and high performance computing support, benchmarking, and reproducibility. The analysis distinguishes hardware demonstrations from simulations and identifies the assumptions and resources that govern comparisons across implementations. Current results do not establish a broad quantum advantage over well matched classical reservoirs. We therefore specify the resource accounting, benchmark standards, and theoretical criteria needed to evaluate claims of quantum advantage.
Reservoir computing (RC) is an effective method for predicting chaotic systems by using a high-dimensional dynamic reservoir with fixed internal weights, while keeping the learning phase linear, which simplifies training and reduces computational complexity compared to fully trained recurrent neural networks (RNNs). Quantum reservoir computing (QRC) uses the exponential growth of Hilbert spaces in quantum systems, allowing for greater information processing, memory capacity, and computational power. However, the original QRC proposal requires coherent injection of inputs multiple times, complicating practical implementation. We present a hybrid quantum-classical approach that implements memory through classical post-processing of quantum measurements. This approach avoids the need for multiple coherent input injections and is evaluated on benchmark tasks, including the chaotic Mackey-Glass time series prediction. We tested our model on two physical platforms: a fully connected Ising model and a Rydberg atom array. The optimized model demonstrates promising predictive capabilities, achieving a higher number of steps compared to previously reported approaches.
This replaces v1, which reported a ten-order shot-noise suppression from a quantum reservoir autoencoder and a two-phase protocol reconstructing unseen inputs at MSE ~ 10^-4. We withdraw both claims: under one consistent evaluation the suppression reverses sign (reset channels degrade reconstruction 144x), the two-phase target is invertible in closed form without training, and a polynomial using no quantum features matches the full model. We trace these to four evaluation artifacts; the fourth also invalidates the companion latent code, which never leaves its initialization. What survives the same controls is a joint scrambling channel: injected on one reservoir half, read out on the other, decoded by a readout calibrated once on training messages. Over 32 realizations the held-out error is below chance in 31-32 (permutation p<10^-4) with a shuffled control at chance. Scrambler T-doping is necessary for one-shot injection, while interleaved injection supplies its own non-Cliffordness; a matched classical scrambler decodes the task better, so the transport is not evidence of a quantum mechanism. The access threshold, recomputed on nested subsystems with right-censoring, gives interleaved medians f* = 0.25-0.33; one-shot injection reaches full detectability in only half the realizations at the largest settings. A trained variational ansatz loses to the fixed untrained circuit by 2.6-4.4x. Three extensions pass the same discipline: streaming requires dissipative reset of the accessible half; the decoder does not transfer across the eight keys tested; a receiver holding the key matches the supervised ceiling from self-simulated pairs. A first hardware run decodes at 0.23 with the shuffled control at chance, while coherent device errors partially open the t=0 control. We close with the controls that separate transport from interpolation.
We study byte-level autoregressive language models whose hidden state is a small quantum register, whose transitions are learned token-conditioned quantum channels, and whose next-byte probabilities come directly from Born-rule measurements without a learned classical vocabulary projection or softmax. In the fixed-width recurrent setting, unitary dynamics cannot contract or reset stored distinctions; experimentally, unitary recurrence loses to its retrained one-token control in every tested variant. Introducing learned dissipation, then token-conditioned dissipative gating, yields useful memory beyond one token. The gated model improves test negative log-likelihood by 0.20 +/- 0.03 nats/byte over retrained parameter-matched one-token controls at context length 12, and by 0.29-0.33 at longer contexts, but never beats retrained two-token controls. A key-value recall benchmark shows a width-dependent threshold: one stored pair meets the pre-registered accuracy bar at 4-5 memory qubits, two pairs at 6-7, and three pairs remain unsolved. This is consistent with, but does not establish, a capacity explanation for the short effective memory horizon. As a secondary result, one-qubit models reach lower stationary memory entropy than a heuristically optimized classical predictive frontier at matched error in 9/9 seeds. On IBM Heron hardware, computational-basis memory-register distributions show signal against a uniform null in 12/12 circuits across prefixes of up to 12 tokens, next-bit probabilities pass pre-registered accuracy bars at four-token prefixes, and bytes are generated autoregressively by sampling physically measured bit probabilities. The tested models are classically simulable and substantially worse than classical n-gram baselines. We claim no computational quantum advantage.
In the field of quantum reservoir computing (QRC), many different computational models and architectures have been proposed. From these models, we identify feedback-based models -- which use a feedback mechanism to re-embed classical measurements from the QRC -- and recurrent models -- which use a multi-register approach with memory and readout qubits -- as the two major competing architectures that have been discussed and validated on hardware. In this paper, we advance upon the recurrent architectures, which employ a two register approach to endow the QRC with a fading memory. While these approaches have been validated on hardware and have demonstrated great real-world performance on noisy-intermediate-scale-quantum (NISQ) quantum processing units (QPUs), the exact mechanism through which the memory capacity arises is not completely understood or fully controllable. With this, we augment the recurrent approaches and present a hardware-realizable mechanism, which we call a tunable partial-SWAP, that allows for the direct control of the rate of memory dissipation from a QRN implemented on a gate-based QPU. The theory behind this mechanism is discussed in terms of a controlled amplitude-damping channel and validation experiments using a randomized short-term memory capacity (STMC) recall benchmark and the NARMA-5 dataset are conducted using simulation and IBM QPUs, respectively.
Quantum reservoir computing (QRC) uses fixed quantum dynamics as a high-dimensional temporal feature map and trains only a lightweight classical readout. QRC is attractive for near-term quantum machine learning, but its performance depends strongly on architecture choices such as input encoding, reservoir depth, entanglement topology, measurement features, state-reset policy, feature construction, and readout regularization. We introduce \method, a simulator-based benchmark that formulates QRC design as constrained black-box architecture search and evaluates whether large language models can act as proposal controllers for this search problem. The benchmark compares five policies under identical evaluation budgets: random search, evolutionary search, Bayesian/TPE optimization, a feedback-based LLM agent, and \hybrid, which combines LLM proposals with memory, mutation, crossover, duplicate avoidance, and exploration. On NARMA10, Mackey-Glass forecasting, and temporal parity, \hybrid{} is the most consistent policy: it ranks first on NARMA10 and temporal parity and second on Mackey-Glass, narrowly behind evolutionary search. Under a 25-evaluation budget and three seeds, \hybrid{} improves over random search on all tasks, including a 23.6\% relative reduction in Mackey-Glass error. The results do not show that LLMs are universal QRC optimizers; rather, they show that generative models can be useful high-level controllers when embedded inside validated, reproducible hybrid search loops.