Challenges
Datasets
Workspaces
Papers
Discussions
Leaderboard
Log inSign up
Challenges
Datasets
Workspaces
Papers
Discussions
Leaderboard
Blog
Job Board
Q3AS

© 2026 Aqora Quantum S.A.S.

TermsPrivacyLegal Notice

arXiv:2611.03845

Entanglement-Aware Quantum Circuit Optimization through Globally Optimal Partitioning

Gregory Morse, Muhammad Al Farizi, Jakab Nádori, László Hajas, Tamás Kozsik, Peter Rakyta

quant-phSubmitted Nov 2, 2026• Announced Nov 4, 2026• 0 citations
arXivHTMLPDF

Abstract

We introduce a scalable framework for globally optimal quantum-circuit partitioning and entanglement-aware block resynthesis. The partitioning problem is formulated as an integer linear program over qubit-feasible convex subcircuits, supporting objectives such as minimizing the number of partitions or a weighted execution cost. The resulting partition selection is exact over the enumerated feasible candidates and their computed costs when the ILP is solved to proven optimality. The selected partitions are resynthesized using an operator-Schmidt rank (OSR) based synthesis technique, which exploits their internal entanglement structure. We compare the end-to-end integer linear programming (ILP)--OSR workflow with BQSKit while identifying their different routing methods, and separately compare local OSR and BQSKit synthesis on matched three- and four-qubit unitaries. On the frozen sample of 772 local unitary targets, OSR produces 741 verified syntheses versus 617 for BQSKit, and matches or improves the CNOT count in all 617 jointly verified pairs. These results demonstrate that globally optimal partition selection combined with entanglement-aware resynthesis improves circuit quality and extends block-based optimization to problem sizes that are impractical for existing optimization workflows.

AI Summary

The paper addresses how to choose bounded-width circuit blocks for resynthesis, and how to synthesize them using their entanglement structure. It combines enumeration of qubit-feasible convex subcircuits with an integer linear program selecting a minimum-cost exact cover. Acyclicity constraints ensure that selected blocks admit a valid execution order. Optimality applies only to the enumerated candidates and their computed costs, when the ILP is solved to proven optimality—not to the final circuit’s minimum possible CNOT count. Local synthesis uses operator-Schmidt spectra of fitted residual unitaries to guide topology-respecting CNOT-structure search. Cut-based bounds, motivated by $\lceil\log_2\mathrm{OSR}\rceil$, and weighted spectral-tail scores prioritize extensions; continuous Hilbert–Schmidt optimization fits single-qubit gates. These are heuristic search rules, not minimum-CNOT guarantees or admissible bounds for all refitted descendants. Acceptance requires numerical unitary verification. On 772 fixed three- and four-qubit targets, with matched topology, $10^{-10}$ decomposition-infidelity tolerance and a 30-second watchdog, OSR verifies 741 syntheses versus BQSKit’s 617. The 617 jointly verified pairs yield 466 CNOT-count wins, 151 ties and no losses; median solver times favor OSR in every reported context. End-to-end benchmarks also report CNOT reductions, but use different routing methods and therefore do not isolate partitioning or synthesis effects. Limitations include exponential worst-case candidate enumeration, substantial runtime outliers, incomplete four-qubit comparisons, and a local benchmark described as calibration rather than held-out evaluation. Larger blocks do not consistently improve routed circuits: across 53 complete-stage circuits, final CNOT totals are 49,621 for three-qubit blocks and 49,300 for four-qubit blocks.

Some signs of AI-generated textAI use disclosedMissing comparisonUnclear novelty
OpenAI GPT-5.5/5.6/6.0/6.1 Sol models, used through the Codex harness in medium and extra-high reasoning modes, assisted with method development, implementation, computational experiments, literature and citation checks, and manuscript preparation.

Error found: The printed partition-enumeration algorithm can repeat an exclusion state indefinitely because it removes the excluded vertex from Y but leaves it eligible for selection in the frontier. The benchmark results omit ILP optimality statuses and gaps, leaving achievement of the claimed conditional partition-selection optimum unverified for those runs.

Written by gpt-6.1-sol from the paper's full text. Automated: it can be wrong.

Comments
29 pages, 11 tables

Related papers

QGo: Scalable Quantum Circuit Optimization Using Automated Synthesis

Xin-Chuan Wu, Davis, Marc Grau, Frederic T. Chong +1 more

The current phase of quantum computing is in the Noisy Intermediate-Scale Quantum (NISQ) era. On NISQ devices, two-qubit gates such as CNOTs are much noisier than single-qubit gates, so it is essential to minimize their count. Quantum circuit synthesis is a process of decomposing an arbitrary unitary into a sequence of quantum gates, and can be used as an optimization tool to produce shorter circuits to improve overall circuit fidelity. However, the time-to-solution of synthesis grows exponentially with the number of qubits. As a result, synthesis is intractable for circuits on a large qubit scale. In this paper, we propose a hierarchical, block-by-block optimization framework, QGo, for quantum circuit optimization. Our approach allows an exponential cost optimization to scale to large circuits. QGo uses a combination of partitioning and synthesis: 1) partition the circuit into a sequence of independent circuit blocks; 2) re-generate and optimize each block using quantum synthesis; and 3) re-compose the final circuit by stitching all the blocks together. We perform our analysis and show the fidelity improvements in three different regimes: small-size circuits on real devices, medium-size circuits on noise simulations, and large-size circuits on analytical models. Using a set of NISQ benchmarks, we show that QGo can reduce the number of CNOT gates by 29.9% on average and up to 50% when compared with industrial compilers such as t|ket>. When executed on the IBM Athens system, shorter depth leads to higher circuit fidelity. We also demonstrate the scalability of our QGo technique to optimize circuits of 60+ qubits. Our technique is the first demonstration of successfully employing and scaling synthesis in the compilation toolchain for large circuits. Overall, our approach is robust for direct incorporation in production compiler toolchains.

arXiv:2012.09835•Dec 17, 2020•52 citations

Optimized Quantum Circuit Partitioning Across Multiple Quantum Processors

Eneet Kaur, Shahrooz Pouryousef, Hassan Shapourian +4 more

This paper addresses the challenge of scaling quantum computing by employing distributed quantum algorithms across multiple processors. As quantum circuits grow in complexity, efficiently managing the entanglement resources required for communication between quantum processing units (QPUs) becomes increasingly critical. We propose novel methods to optimize entanglement consumption while accounting for network constraints. Our approach leverages graph partitioning techniques to optimize both qubit teleportation and gate teleportation, minimizing the number of Einstein-Podolsky-Rosen (EPR) pairs required to execute general quantum circuits. Additionally, we formulate an integer linear program (ILP) to further reduce entanglement requirements by mapping the logical resources of partitioned circuits to the physical constraints of the underlying quantum network. Finally, we analyze the entanglement cost for implementing the Quantum Fourier Transform (QFT) across multiple QPUs, employing an approach that exploits the circuit's structure to minimize total entanglement consumption. These contributions provide a pathway to scalable and resource-efficient distributed quantum computing.

IEEE Transactions on Quantum Engineering•Jan 1, 2025•9 citations

Reoptimization of Quantum Circuits via Hierarchical Synthesis

Xin-Chuan Wu, Marc Grau Davis, Frederic T. Chong +1 more

The current phase of quantum computing is in the Noisy Intermediate-Scale Quantum (NISQ) era. On NISQ devices, two-qubit gates such as CNOTs are much noisier than single-qubit gates, so it is essential to minimize their count. Quantum circuit synthesis is a process of decomposing an arbitrary unitary into a sequence of quantum gates, and can be used as an optimization tool to produce shorter circuits to improve overall circuit fidelity. However, the time-to-solution of synthesis grows exponentially with the number of qubits. As a result, synthesis is intractable for circuits on a large qubit scale. In this paper, we propose a hierarchical, block-by-block opti-mization framework, QGo, for quantum circuit optimization. Our approach allows an exponential cost optimization to scale to large circuits. QGo uses a combination of partitioning and synthesis: 1) partition the circuit into a sequence of independent circuit blocks; 2) re-generate and optimize each block using quantum synthesis; and 3) re-compose the final circuit by stitching all the blocks together. We perform our analysis and show the fidelity improvements in three different regimes: small-size circuits on real devices, medium-size circuits on noisy simulations, and large-size circuits on analytical models. Our technique can be applied after existing optimizations to achieve higher circuit fidelity. Using a set of NISQ benchmarks, we show that QGo can reduce the number of CNOT gates by 29.9% on average and up to 50% when compared with industrial compiler optimizations such as t|ket). When executed on the IBM Athens system, shorter depth leads to higher circuit fidelity. We also demonstrate the scalability of our QGo technique to optimize circuits of 60+ qubits, Our technique is the first demonstration of successfully employing and scaling synthesis in the compilation tool chain for large circuits. Overall, our approach is robust for direct incorporation in production compiler toolchains to further improve the circuit fidelity.

IEEE International Conference on Rebooting Computing (ICRC)•Nov 1, 2021•23 citations

Tackling the Qubit Mapping Problem with Permutation-Aware Synthesis

Ji Liu, Ed Younis, Mathias Weiden +3 more

We propose a scalable, hierarchical qubit mapping and routing algorithm that harnesses the power of circuit synthesis. First, we decompose large circuits into subcircuits small enough to be directly resynthesized. For each block, we pre-synthesize them for all permutations of its input and output qubits. Following this offline step, we employ a permutation-aware, block-based generalization of the popular SABRE mapping algorithm. This mapping step stitches together blocks by choosing an input-output permutation that minimizes intrablock gate count and required inter-block communication (SWAP and bridge gates). Our approach has a twofold advantage: 1) circuit synthesis may eliminate more two-qubit gates than other optimizing compilers; 2) considering all permutations of input and output qubits eliminates communication operations transparently. In contrast, other mapping algorithms can only introduce communication operations. We show that we can produce better-quality circuits than commercial compilers: shorter by up to 68% (18% on average) fewer gates than Qiskit, up to 36% (9% on average) fewer gates than Tket. We outperform BQSkit, a permutation-unaware, synthesis-based compiler, by up to 67% (21% on average) fewer gates. We also exceed experimental optimal mappers such as OLSQ in quality (10.7% shorter circuits) and time to solution. Our scalable, heuristic approach can be seamlessly integrated into any quantum circuit compiler or optimization infrastructure, and it applies well to any qubit technology, such as superconducting and trapped ions.

arXiv:2305.02939•IEEE International Conference on Quantum Computing and Engineering (QCE)•Sep 17, 2023•37 citations

Provably Optimal Quantum Circuits with Mixed-Integer Programming

Harsha Nagarajan, Zsolt Szabó

We present a depth-aware optimization framework for quantum circuit compilation that unifies provable optimality with scalable heuristics. For exact synthesis of a target unitary, we formulate a mixed-integer linear program (MILP) that linearly handles global-phase equivalence and uses explicit parallel scheduling variables to certify depth-optimal solutions for small-to-medium circuits. Domain-specific valid constraints, including identity ordering, commuting-gate pruning, short-sequence redundancy cuts, and Hermitian-conjugate linkages, significantly accelerate branch-and-bound, yielding speedups up to 43x on standard benchmarks. The framework supports hardware-aware objectives, enabling fault-tolerant (e.g. T-count) and NISQ-era (e.g. entangling gates) devices. For approximate synthesis, we propose 3 objectives: (i) exact, but non-convex, phase-invariant fidelity maximization; (ii) a linear surrogate that maximizes the real trace overlap, yielding a tight lower bound to fidelity; and (iii) a convex quadratic function that minimizes the circuit's Frobenius error. To scale beyond exact MILP, we propose a novel rolling-horizon optimization (RHO) that rolls primarily in time, caps the active-qubits, and enforces per-qubit closure while globally optimizing windowed segments. This preserves local context, reduces the Hilbert-space dimension, and enables iterative improvements without ancillas. On a 142-gate seed circuit, RHO yields 116 gates, an 18.3% reduction from the seed, while avoiding the trade-off between myopic passes and long run times. Empirically, our exact compilation framework achieves certified depth-optimal circuits on standard targets, high-fidelity Fibonacci-anyon weaves, and a 36% gate-count reduction on multi-body parity circuits. All methods are in the open-source QuantumCircuitOpt, providing a single framework that bridges exact certification and scalable synthesis.

arXiv:2510.00649•Oct 1, 2025•5 citations

Entanglement-Capability-Aware Finite-Bond Circuit Cutting for Distributed Quantum Computing

Weidong Liu, Qingqing Jiang, Yufu Liu +4 more

Distributed quantum computing distributes workloads that exceed a single processor's capacity across resource-constrained components. Circuit cutting enables this decomposition, but its effectiveness depends on partition choice. We present SACE, a Simulation-Aware Circuit Partitioning framework with Entanglement-Capability-Aware Screening. EP-Cost screens partitions using a parameter-dependent entanglement-capability proxy and native implementation cost. Under reconstruction constraints, finite-bond refinement selects the candidate that minimizes the worst-subcircuit discrepancy between normalized MPS simulations at bond dimensions $χ$ and $2χ$. Across 14 exhaustively enumerated small-scale candidate spaces, EP-Cost-guided genetic algorithm selections attain co-optimal native quasiprobability decomposition (QPD) overhead, with a median within-circuit Spearman correlation of 0.9965 between EP-Cost and QPD overhead. At matched cut counts, EP-Cost reduces QPD overhead by $25.20\times$ and $16.89\times$ for 8- and 9-qubit QFT circuits, respectively, and improves sampled reconstruction. On 12-qubit approximate QFT, deterministic reconstruction reduces Pauli-observable mean absolute error in all six low-bond conditions and total-variation distance in five. With five cuts and 7,776 QPD terms, SACE is best on all four reconstruction metrics in four of six conditions. At 24 qubits, SACE improves exact-local QFT accuracy in all 30 instance-bond-dimension conditions against each EP-Cost-only, Kernighan-Lin, and spectral baseline. The 50-qubit QFT comparisons remain favorable as reference convergence improves. Weighted QAOA circuits provide complementary, more bond-sensitive evidence. Refinement benefits persist across three search methods; larger candidate panels improve the optimized proxy without guaranteeing lower exact-local error.

quant-pharXiv:2610.14580•Oct 13, 2026•0 citations
We introduce a scalable framework for globally optimal quantum-circuit partitioning and entanglement-aware block resynthesis. The partitioning problem is formulated as an integer linear program over qubit-feasible convex subcircuits, supporting objectives such as minimizing the number of partitions or a weighted execution cost. The resulting partition selection is exact over the enumerated feasible candidates and their computed costs when the ILP is solved to proven optimality. The selected partitions are resynthesized using an operator-Schmidt rank (OSR) based synthesis technique, which exploits their internal entanglement structure. We compare the end-to-end integer linear programming (ILP)--OSR workflow with BQSKit while identifying their different routing methods, and separately compare local OSR and BQSKit synthesis on matched three- and four-qubit unitaries. On the frozen sample of 772 local unitary targets, OSR produces 741 verified syntheses versus 617 for BQSKit, and matches or improves the CNOT count in all 617 jointly verified pairs. These results demonstrate that globally optimal partition selection combined with entanglement-aware resynthesis improves circuit quality and extends block-based optimization to problem sizes that are impractical for existing optimization workflows.
The paper addresses how to choose bounded-width circuit blocks for resynthesis, and how to synthesize them using their entanglement structure. It combines enumeration of qubit-feasible convex subcircuits with an integer linear program selecting a minimum-cost exact cover. Acyclicity constraints ensure that selected blocks admit a valid execution order. Optimality applies only to the enumerated candidates and their computed costs, when the ILP is solved to proven optimality—not to the final circuit’s minimum possible CNOT count.
Local synthesis uses operator-Schmidt spectra of fitted residual unitaries to guide topology-respecting CNOT-structure search. Cut-based bounds, motivated by ⌈log⁡2OSR⌉\lceil\log_2\mathrm{OSR}\rceil⌈log2​OSR⌉, and weighted spectral-tail scores prioritize extensions; continuous Hilbert–Schmidt optimization fits single-qubit gates. These are heuristic search rules, not minimum-CNOT guarantees or admissible bounds for all refitted descendants. Acceptance requires numerical unitary verification.
On 772 fixed three- and four-qubit targets, with matched topology, 10−1010^{-10}10−10 decomposition-infidelity tolerance and a 30-second watchdog, OSR verifies 741 syntheses versus BQSKit’s 617. The 617 jointly verified pairs yield 466 CNOT-count wins, 151 ties and no losses; median solver times favor OSR in every reported context. End-to-end benchmarks also report CNOT reductions, but use different routing methods and therefore do not isolate partitioning or synthesis effects.
Limitations include exponential worst-case candidate enumeration, substantial runtime outliers, incomplete four-qubit comparisons, and a local benchmark described as calibration rather than held-out evaluation. Larger blocks do not consistently improve routed circuits: across 53 complete-stage circuits, final CNOT totals are 49,621 for three-qubit blocks and 49,300 for four-qubit blocks.
The current phase of quantum computing is in the Noisy Intermediate-Scale Quantum (NISQ) era. On NISQ devices, two-qubit gates such as CNOTs are much noisier than single-qubit gates, so it is essential to minimize their count. Quantum circuit synthesis is a process of decomposing an arbitrary unitary into a sequence of quantum gates, and can be used as an optimization tool to produce shorter circuits to improve overall circuit fidelity. However, the time-to-solution of synthesis grows exponentially with the number of qubits. As a result, synthesis is intractable for circuits on a large qubit scale. In this paper, we propose a hierarchical, block-by-block optimization framework, QGo, for quantum circuit optimization. Our approach allows an exponential cost optimization to scale to large circuits. QGo uses a combination of partitioning and synthesis: 1) partition the circuit into a sequence of independent circuit blocks; 2) re-generate and optimize each block using quantum synthesis; and 3) re-compose the final circuit by stitching all the blocks together. We perform our analysis and show the fidelity improvements in three different regimes: small-size circuits on real devices, medium-size circuits on noise simulations, and large-size circuits on analytical models. Using a set of NISQ benchmarks, we show that QGo can reduce the number of CNOT gates by 29.9% on average and up to 50% when compared with industrial compilers such as t|ket>. When executed on the IBM Athens system, shorter depth leads to higher circuit fidelity. We also demonstrate the scalability of our QGo technique to optimize circuits of 60+ qubits. Our technique is the first demonstration of successfully employing and scaling synthesis in the compilation toolchain for large circuits. Overall, our approach is robust for direct incorporation in production compiler toolchains.
This paper addresses the challenge of scaling quantum computing by employing distributed quantum algorithms across multiple processors. As quantum circuits grow in complexity, efficiently managing the entanglement resources required for communication between quantum processing units (QPUs) becomes increasingly critical. We propose novel methods to optimize entanglement consumption while accounting for network constraints. Our approach leverages graph partitioning techniques to optimize both qubit teleportation and gate teleportation, minimizing the number of Einstein-Podolsky-Rosen (EPR) pairs required to execute general quantum circuits. Additionally, we formulate an integer linear program (ILP) to further reduce entanglement requirements by mapping the logical resources of partitioned circuits to the physical constraints of the underlying quantum network. Finally, we analyze the entanglement cost for implementing the Quantum Fourier Transform (QFT) across multiple QPUs, employing an approach that exploits the circuit's structure to minimize total entanglement consumption. These contributions provide a pathway to scalable and resource-efficient distributed quantum computing.
The current phase of quantum computing is in the Noisy Intermediate-Scale Quantum (NISQ) era. On NISQ devices, two-qubit gates such as CNOTs are much noisier than single-qubit gates, so it is essential to minimize their count. Quantum circuit synthesis is a process of decomposing an arbitrary unitary into a sequence of quantum gates, and can be used as an optimization tool to produce shorter circuits to improve overall circuit fidelity. However, the time-to-solution of synthesis grows exponentially with the number of qubits. As a result, synthesis is intractable for circuits on a large qubit scale. In this paper, we propose a hierarchical, block-by-block opti-mization framework, QGo, for quantum circuit optimization. Our approach allows an exponential cost optimization to scale to large circuits. QGo uses a combination of partitioning and synthesis: 1) partition the circuit into a sequence of independent circuit blocks; 2) re-generate and optimize each block using quantum synthesis; and 3) re-compose the final circuit by stitching all the blocks together. We perform our analysis and show the fidelity improvements in three different regimes: small-size circuits on real devices, medium-size circuits on noisy simulations, and large-size circuits on analytical models. Our technique can be applied after existing optimizations to achieve higher circuit fidelity. Using a set of NISQ benchmarks, we show that QGo can reduce the number of CNOT gates by 29.9% on average and up to 50% when compared with industrial compiler optimizations such as t|ket). When executed on the IBM Athens system, shorter depth leads to higher circuit fidelity. We also demonstrate the scalability of our QGo technique to optimize circuits of 60+ qubits, Our technique is the first demonstration of successfully employing and scaling synthesis in the compilation tool chain for large circuits. Overall, our approach is robust for direct incorporation in production compiler toolchains to further improve the circuit fidelity.
We propose a scalable, hierarchical qubit mapping and routing algorithm that harnesses the power of circuit synthesis. First, we decompose large circuits into subcircuits small enough to be directly resynthesized. For each block, we pre-synthesize them for all permutations of its input and output qubits. Following this offline step, we employ a permutation-aware, block-based generalization of the popular SABRE mapping algorithm. This mapping step stitches together blocks by choosing an input-output permutation that minimizes intrablock gate count and required inter-block communication (SWAP and bridge gates). Our approach has a twofold advantage: 1) circuit synthesis may eliminate more two-qubit gates than other optimizing compilers; 2) considering all permutations of input and output qubits eliminates communication operations transparently. In contrast, other mapping algorithms can only introduce communication operations. We show that we can produce better-quality circuits than commercial compilers: shorter by up to 68% (18% on average) fewer gates than Qiskit, up to 36% (9% on average) fewer gates than Tket. We outperform BQSkit, a permutation-unaware, synthesis-based compiler, by up to 67% (21% on average) fewer gates. We also exceed experimental optimal mappers such as OLSQ in quality (10.7% shorter circuits) and time to solution. Our scalable, heuristic approach can be seamlessly integrated into any quantum circuit compiler or optimization infrastructure, and it applies well to any qubit technology, such as superconducting and trapped ions.
We present a depth-aware optimization framework for quantum circuit compilation that unifies provable optimality with scalable heuristics. For exact synthesis of a target unitary, we formulate a mixed-integer linear program (MILP) that linearly handles global-phase equivalence and uses explicit parallel scheduling variables to certify depth-optimal solutions for small-to-medium circuits. Domain-specific valid constraints, including identity ordering, commuting-gate pruning, short-sequence redundancy cuts, and Hermitian-conjugate linkages, significantly accelerate branch-and-bound, yielding speedups up to 43x on standard benchmarks. The framework supports hardware-aware objectives, enabling fault-tolerant (e.g. T-count) and NISQ-era (e.g. entangling gates) devices. For approximate synthesis, we propose 3 objectives: (i) exact, but non-convex, phase-invariant fidelity maximization; (ii) a linear surrogate that maximizes the real trace overlap, yielding a tight lower bound to fidelity; and (iii) a convex quadratic function that minimizes the circuit's Frobenius error. To scale beyond exact MILP, we propose a novel rolling-horizon optimization (RHO) that rolls primarily in time, caps the active-qubits, and enforces per-qubit closure while globally optimizing windowed segments. This preserves local context, reduces the Hilbert-space dimension, and enables iterative improvements without ancillas. On a 142-gate seed circuit, RHO yields 116 gates, an 18.3% reduction from the seed, while avoiding the trade-off between myopic passes and long run times. Empirically, our exact compilation framework achieves certified depth-optimal circuits on standard targets, high-fidelity Fibonacci-anyon weaves, and a 36% gate-count reduction on multi-body parity circuits. All methods are in the open-source QuantumCircuitOpt, providing a single framework that bridges exact certification and scalable synthesis.
Distributed quantum computing distributes workloads that exceed a single processor's capacity across resource-constrained components. Circuit cutting enables this decomposition, but its effectiveness depends on partition choice. We present SACE, a Simulation-Aware Circuit Partitioning framework with Entanglement-Capability-Aware Screening. EP-Cost screens partitions using a parameter-dependent entanglement-capability proxy and native implementation cost. Under reconstruction constraints, finite-bond refinement selects the candidate that minimizes the worst-subcircuit discrepancy between normalized MPS simulations at bond dimensions χχχ and 2χ2χ2χ. Across 14 exhaustively enumerated small-scale candidate spaces, EP-Cost-guided genetic algorithm selections attain co-optimal native quasiprobability decomposition (QPD) overhead, with a median within-circuit Spearman correlation of 0.9965 between EP-Cost and QPD overhead. At matched cut counts, EP-Cost reduces QPD overhead by 25.20×25.20\times25.20× and 16.89×16.89\times16.89× for 8- and 9-qubit QFT circuits, respectively, and improves sampled reconstruction. On 12-qubit approximate QFT, deterministic reconstruction reduces Pauli-observable mean absolute error in all six low-bond conditions and total-variation distance in five. With five cuts and 7,776 QPD terms, SACE is best on all four reconstruction metrics in four of six conditions. At 24 qubits, SACE improves exact-local QFT accuracy in all 30 instance-bond-dimension conditions against each EP-Cost-only, Kernighan-Lin, and spectral baseline. The 50-qubit QFT comparisons remain favorable as reference convergence improves. Weighted QAOA circuits provide complementary, more bond-sensitive evidence. Refinement benefits persist across three search methods; larger candidate panels improve the optimized proxy without guaranteeing lower exact-local error.