Programmable Dissipation via Partial Quantum Error Correction
Sameer Dambal, Michael AD Taylor, Yu Zhang · 2026
Read Programmable Dissipation via Partial Quantum Error Correction by Sameer Dambal, Michael AD Taylor, Yu Zhang (2026) in the Tower of Babel, a free online library of public-domain works preserved in clean, searchable text. No sign-up. The full text is served as a plain-text file at a predictable URL, where text is present: https://antarctic-labs.com/tower-of-babel/arxiv-2605-30217.txt.
Introduction
Sameer Dambal, Michael AD Taylor and Yu Zhang show how partial quantum error correction turns logical noise into a programmable resource for open simulations. Noise usually fights quantum information, yet open systems need dissipation. Treating fault-tolerant rounds as programmable units, decoder and recovery randomization generate controllable logical channels whose mixtures realize Kraus-channel mixing. Target dissipators can thus be compiled into logical dynamics without extra ancilla qubits for the bath. An accuracy criterion ties code distance to keeping uncontrolled errors a small fraction of intended dissipation.
Catalog record
| CREATOR | Sameer Dambal, Michael AD Taylor, Yu Zhang |
|---|---|
| YEAR | 2026 |
| COLLECTION | PAPERS |
| FORMAT | TXT |
| FILE SIZE | 32174 |
| LICENSE | http://creativecommons.org/licenses/by/4.0/ |
| RIGHTS | CC BY |
| SOURCE | arXiv |
Source: https://arxiv.org/abs/2605.30217
Browse the full library · About the Tower of Babel
Full text
The opening of the work, from the library's hosted text:
Programmable Dissipation via Partial Quantum Error Correction Sameer Dambal, Michael AD Taylor, Yu Zhang Abstract Noise is typically treated as the adversary of quantum information processing. For open quantum dynamics, however, dissipation is part of the target physics, creating a tension with fault-tolerant architectures designed to suppress decoherence. Here we show that logical noise can instead be turned into a calibrated resource. We treat the error-correction cycle as a programmable primitive: one fault-tolerant round induces a logical completely positive trace-preserving map, and decoder/recovery randomization generates a controllable family of logical channels whose convex mixtures realize Kraus-channel mixing. This enables direct compilation of target dissipators into effective logical dynamics without explicit ancilla qubits for encoding the bath degree of freedoms. We derive an accuracy criterion for multi-step simulation in which the code distance is chosen so that uncontrolled logical errors remain a small fraction of the intended dissipation per step, rather than being driven below an arbitrarily small closed-system tolerance. Partial quantum error correction thus repurposes fault-tolerant structure to sculpt dissipation, offering a resource-efficient route to quantum simulation of open quantum systems. Introduction . Noise has traditionally been the adversary of quantum computation. In the noisy intermediate-scale quantum (NISQ) era, this view has motivated shallow, noise-resilient circuit reduction methods 2 ; 47 and a broad class of error-mitigation techniques 7 ; 41 designed to extract useful signals before decoherence overwhelms the computation. As the field moves toward fault-tolerant quantum computation, the emphasis shifts from tolerating noise to actively suppressing it: quantum error correction (QEC) encodes logical information into redundant physical degrees of freedom, repeatedly extracts syndromes, and uses recovery operations to make the logical channel as close as possible to the intended unitary evolution 37 ; 38 ; 22 ; 8 ; 26 ; 13 ; 17 ; 40 ; 43 ; 3 ; 25 ; 1 . This architecture is essential for high-fidelity quantum algorithms, but it is not naturally matched to many problems in quantum science. Real quantum systems are rarely perfectly isolated: dissipation and dephasing are not merely imperfections but often constitute the dynamics to be predicted, as in quantum transport, spin relaxation, non-equilibrium materials, quantum optical systems, and beyond 6 ; 18 ; 28 ; 19 ; 30 ; 31 ; 16 ; 29 ; 48 ; 9 . This mismatch creates a twofold overhead for simulating open quantum systems on fault-tolerant devices. First, many physical qubits are consumed to construct logical qubits. Second, the desired non-unitary reduced dynamics must be embedded back into a larger unitary evolution, for example through a Stinespring dilation, stochastic unraveling, postselection, block encoding, or related ancilla-based constructions 23 ; 39 ; 32 ; 45 ; 4 ; 33 ; 15 ; 36 ; 21 ; 44 ; 12 . Here we take a different route: we propose partial QEC as a logical dissipative primitive. A syndrome-extraction cycle is a physical completely positive trace-preserving (CPTP) process once measurement outcomes are averaged over and the classical record is discarded. Encoding and decoding convert this physical process into an effective logical CPTP map. Moreover, if the recovery operation, decoder tie breaking, logical-frame update, or a small recovery gadget is sampled from a classical distribution, the resulting logical channel is a convex combination of implementable logical channels. The partial QEC proposed in this work uses this freedom not merely to preserve information, as in approximate or noise-adapted QEC 27 ; 5 ; 24 , but to synthesize a target dissipative map directly at the logical level. We propose two complementary strategies to achieve the goal. Strategy A is model-aware: the calibrated physical noise, syndrome circuit, decoder, and randomized recovery library are propagated to a family of logical maps 41 . The target dissipator is then fitted within the convex hull of that family. Logical noise aligned with the target dissipator is no longer counted as failure, so the code distance is set by the residual wrong-channel errors, fitting error and model mismatch. Strategy B is post-correction programming: the physical noise is first removed by conventional QEC, and a logical CPTP map is then sampled from an additional programming library, like ancilla-enabled corrected logical operations 35 . This route is model-agnostic with respect to the physical-to-logical mapping, and it can implement nonunitary logical dynamics without allocating explicit bath registers for a Stinespring dilation. Strategy B therefore does not reduce the code distance required for the correction stage, but it can reduce the channel-implementation overhead relative to dilation-based simulation. Results and Discussion . For a system \mathcal{S} interacting with an environment \mathcal{E} , the total Hamiltonian H_{T} is given by H_{T}=H_{\mathcal{S}}+H_{\mathcal{E}}+H_{\mathcal{SE}} . With \hbar=1 , the dynamics of the reduced system can be obtained by tracing out the environmental DoF from the unitary dynamics of the entire system, i.e., i\frac{\partial\rho_{\mathcal{S}}(t)}{\partial t}=\text{Tr}_{\mathcal{E}}[H_{T},\rho_{T}]. (1) For an initially factorized preparation \rho_{T}(0)=\rho_{\mathcal{S}}(0)\otimes\rho_{\mathcal{E}} , the dynamics of the system is given by \rho_{\mathcal{S}}(t)=\operatorname{Tr}_{\mathcal{E}}\left[U(t)\bigl(\rho_{\mathcal{S}}(0)\otimes\rho_{\mathcal{E}}\bigr)U^{\dagger}(t)\right]=\mathcal{E}_{\rm tar}^{(t)}(\rho_{\mathcal{S}}(0)), (2) where U=e^{-iH_{T}t} . For a fixed preparation of the environment, this map is completely positive and trace-preserving (CPTP) and admits a Kraus representation 23 , \rho_{\mathcal{S}}(t)=\sum_{i}K_{i}\rho_{\mathcal{S}}K_{i}^{\dagger},\quad\sum_{i}K_{i}^{\dagger}K_{i}=\mathbb{I}. (3) For example, for the widely used Lindbladian dynamics, its corresponding Kraus formalism is K_{0}=\mathbb{I}-(iH+\frac{1}{2}\sum_{i}L^{\dagger}_{i}L_{i})\Delta t,\quad K_{i}=\sqrt{\Delta t}L_{i} . In fault-tolerant architectures, however, the elementary cycle is not the target dissipative dynamics \mathcal{E}_{\mathrm{tar}}^{(\tau)} but a quantum error correction (QEC) 14 round: syndrome extraction ( \mathcal{S} ), decoding ( \mathcal{D} ), and conditional recovery \mathcal{R} . This QEC round naturally defines a CPTP map on the encoded degrees of freedom (DoF) when one averages over all syndrome outcomes. The central idea of partial QEC is to use this map not only to suppress errors, but also to shape the effective logical dynamics so that inevitable logical noise becomes the intended dissipation. To formalize this perspective, let \mathcal{C}_{d} denote an ideal encoding map into a distance- d code space, and let a single fault-tolerant round (including physical noise, syndrome extraction, and conditional recovery) define an instrument \{\mathcal{M}_{s}\} followed by recovery channels \{\mathcal{A}_{s}\} . The syndrome-averaged channel on the encoded block is \mathcal{Q}_{d}=\sum_{s}\mathcal{A}_{s}\circ\mathcal{M}_{s}, (4) where \circ denotes composition of maps (channels); for instance, (\mathcal{D}\circ\mathcal{U})(\rho)=\mathcal{D}(U\rho U^{\dagger}) . The induced logical channel (in an ideal decode picture) is \mathcal{B}_{d}=\mathcal{D}_{d}\circ\mathcal{Q}_{d}\circ\mathcal{C}_{d}, (5) where \mathcal{D}_{d} is an ideal decoding map back to the logical Hilbert space 1 1 1 Equivalently one may stay entirely within the code space and use logical Pauli/frame tracking; the abstract channel \mathcal{B}_{d} captures the logical input–output relation after discarding classical records. . However, we will show (prove) that one QEC round can fundamentally generate a CPTP logical channel. Lemma 1 (One QEC round induces a CPTP logical channel) . Let D denote the data block (the encoded physical qubits) and A the syndrome ancillas. Assume that in each QEC round: (i) ancillas are prepared in a fixed state \omega_{A} (freshly prepared or reset), (ii) a fixed noisy syndrome-extraction circuit is applied (modeled as a CPTP channel on DA ), (iii) ancillas are measured with outcomes s and we do not postselect (we sum over all s ), (iv) conditioned on s , a recovery channel \mathcal{A}_{s} (TP) is applied to D , (v) ancillas and classical records are discarded. Then the syndrome-averaged map (Eq. ( 4 )) on the data block, i.e., \mathcal{Q}_{d}(\rho_{D})=\sum_{s}\mathcal{A}_{s}\left(\mathcal{M}_{s}(\rho_{D})\right) ,