When Close Enough Is Not Enough: Autoregressive Drift in Quantum Circuit Synthesis

• URL: http://arxiv.org/abs/2607.12780v1
• Date: Tue, 14 Jul 2026 13:56:39 GMT
• Title: When Close Enough Is Not Enough: Autoregressive Drift in Quantum Circuit Synthesis
• Abstract summary: Quantum circuit optimization for fault-tolerant computing requires exact functional equivalence while minimizing expensive non-Clifford resources such as T gates.<n>We study this problem using a compact 44.8M- parameter encoder-decoder transformer with structured circuit tokenization.<n>On parameterized circuits, a hybrid approach -- structure from the transformer, angles from classical optimization -- achieves median fidelity 1.000 on 3-6 qubit circuits.
• Score: 0.0
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Quantum circuit optimization for fault-tolerant computing requires exact functional equivalence while minimizing expensive non-Clifford resources such as T gates. We study this problem using a compact 44.8M-parameter encoder-decoder transformer with structured circuit tokenization, evaluating on parameterized circuits (2-6 qubits) and Clifford+T circuits (3-6 qubits). On parameterized circuits, a hybrid approach -- structure from the transformer, angles from classical optimization -- achieves median fidelity 1.000 on 3-6 qubit circuits. On Clifford+T circuits, where all gates are discrete and no post-processing is possible, the model learns valid syntax and accurate T-Count statistics, yet exact equivalence degrades sharply with target length -- from 88% on circuits with <=9 gates to near zero beyond 26 gates. We trace this failure to autoregressive drift: early-token divergence cascading irrecoverably through left-to-right decoding. Two levers partially mitigate the drift: inference-time strategies that generate multiple candidates and select via equivalence verification raise exact-match rates from 7% to 22.5%, while scaling training data by 2.5x pushes them to 39.5%. Yet the degradation with target length persists -- even with more data, exact equivalence drops from 94% on short circuits to under 4% beyond 26 gates. The contrast between settings is our central finding: when approximate outputs can be rescued by post-processing, the transformer succeeds; when exact discrete correctness is required, autoregressive drift limits reliability, with both inference-time search and data scaling as effective levers while training-side fine-tuning and model-level diversification are not.

Related papers

• CircuitsDNA: Discovering Unconventional Multi-Accuracy Arithmetic Circuits via Evolutionary Synthesis [1.10042698101905]
  AI workloads increasingly require arithmetic units that can trade computational accuracy for efficiency on demand.<n>This work introduces CircuitsDNA, an evolutionary framework that automatically evolves accuracy-configurable arithmetic circuits.<n>It integrates three key features: 1) multi-threshold verifiability miter to enforce mode-specific accuracy requirements, 2) resource-limited verifiability-driven search to reduce verification overhead without sacrificing correctness, and 3) feedback-driven adaptive mutation to prioritize effective structural modifications.
  arXiv  Detail & Related papers  (2026-09-01T18:03:46Z)
• Efficient Circuit Transpilation of Commuting Gates on 2D Grids [0.21485350418225238]
  We propose a transpilation scheme for circuits consisting of blocks of commuting two-qubit gates on two-dimensional lattices.<n>We benchmark our approach on QAOA instances for Maximum Cut (MC) on Random Regular graphs and Maximum Independent Set (MIS) on Erds-Rényi graphs.<n>Compared to standard methods, we reduce circuit depth and gate count by about a factor of two, enabling experiments with up to $80$ qubits and improving approximation ratios by up to $6.6%$ for MC and $9.3%$ for MIS.
  arXiv  Detail & Related papers  (2026-07-11T12:09:27Z)
• Constrained Decoding for Diffusion Language Models via Efficient Inference over Finite Automata [57.27430779838529]
  We present an exact and tractable algorithm for sampling from the constrained mean-field posterior under any constraint expressible as a finite automaton.<n>The approach guarantees constraint satisfaction by construction, supports both greedy and sampling-based decoding, and is compatible with parallel and block-wise decoding.<n> Empirical evaluations on Dream-7B and LLaDA-8 show substantial accuracy gains across various tasks.
  arXiv  Detail & Related papers  (2026-07-08T05:48:57Z)
• Equivariant Reinforcement Learning for Clifford Quantum Circuit Synthesis [2.5234156040689233]
  We consider the problem of Clifford quantum circuits for devices with all-to-all qubit connectivity.<n>We introduce a novel neural network architecture that is equivariant to qubit relabelings of the symplectic matrix representation.
  arXiv  Detail & Related papers  (2026-05-11T17:49:28Z)
• BPDQ: Bit-Plane Decomposition Quantization on a Variable Grid for Large Language Models [56.504879072674015]
  We propose Bit-Plane Decomposition Quantization (BPDQ), which constructs a variable quantization grid via bit-planes and scalar coefficients.<n>BPDQ enables serving Qwen2.5-72B on a single GTX 3090 with 83.85% GSM8K accuracy (vs. 90.83% at 16-bit)
  arXiv  Detail & Related papers  (2026-02-04T02:54:37Z)
• Noise-Adaptive Quantum Circuit Mapping for Multi-Chip NISQ Systems via Deep Reinforcement Learning [0.0]
  We present DeepQMap, a deep reinforcement learning framework that integrates a bidirectional Long Short-Term Memory based Dynamic Noise Adaptation network.<n>Our method continuously adapts to hardware dynamics through learned temporal representations of quantum system behavior.<n>DeepQMap achieves mean circuit fidelity of $0.920 pm 0.023$, representing a statistically significant 49.3% improvement over state-of-the-art QUBO methods.
  arXiv  Detail & Related papers  (2025-11-22T14:27:55Z)
• Above 99.9% Fidelity Single-Qubit Gates, Two-Qubit Gates, and Readout in a Single Superconducting Quantum Device [58.154405222706146]
  tuning of qubit-coupler coupling strengths in a superconducting circuit with two transmon qubits coupled via a tunable coupler enables high-fidelity single- and two-qubit gates.<n>We achieve a 40h-averaged CZ gate fidelity of 99.93%, simultaneous single-qubit gate fidelities of 99.98%, and readout fidelities over 99.94% in a single device.
  arXiv  Detail & Related papers  (2025-08-22T14:49:47Z)
• Fast, tunable, high fidelity cZ-gates between superconducting qubits with parametric microwave control of ZZ-coupling [0.0]
  We present a highly flexible parametric coupling scheme with superconducting qubits. Our fully integrated, 2D on-chip coupler design is only weakly flux tunable. benchmarking reveals that the parametric SWAP c$Z$ gate achieves an average fidelity of $99.44pm 0.09$% in a gate duration of 70ns.
  arXiv  Detail & Related papers  (2023-05-04T15:14:56Z)
• Accurate methods for the analysis of strong-drive effects in parametric gates [94.70553167084388]
  We show how to efficiently extract gate parameters using exact numerics and a perturbative analytical approach. We identify optimal regimes of operation for different types of gates including $i$SWAP, controlled-Z, and CNOT.
  arXiv  Detail & Related papers  (2021-07-06T02:02:54Z)
• Clifford Circuit Optimization with Templates and Symbolic Pauli Gates [11.978356827088595]
  The Clifford group is a finite subgroup of the unitary group generated by the Hadamard, the CNOT, and the Phase gates. Here we consider the problem of finding a short quantum circuit implementing a given Clifford group element. Our methods aim to minimize the entangling gate count assuming all-to-all qubit connectivity.
  arXiv  Detail & Related papers  (2021-05-05T19:18:35Z)
• On the realistic worst case analysis of quantum arithmetic circuits [69.43216268165402]
  We show that commonly held intuitions when designing quantum circuits can be misleading. We show that reducing the T-count can increase the total depth. We illustrate our method on addition and multiplication circuits using ripple-carry.
  arXiv  Detail & Related papers  (2021-01-12T21:36:16Z)
• The cost of universality: A comparative study of the overhead of state distillation and code switching with color codes [63.62764375279861]
  We compare two leading FT implementations of the T gate in 2D color codes under circuit noise. We find a circuit noise threshold of 0.07(1)% for the T gate via code switching, almost an order of magnitude below that achievable by state distillation in the same setting.
  arXiv  Detail & Related papers  (2021-01-06T19:00:01Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.