Pith. sign in

REVIEW 3 major objections 4 minor 93 references

Position: Quantum Program Generation Must Prioritize Validity Over Probabilistic Scaling

T0 review · 3 major / 4 minor · reviewed 2026-08-02 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗

Pith's one-line read Quantum program generation must prioritize validity over scale: post-hoc filtering cannot close the exponential gap.

desk verdict A well-argued position for verifier-centric quantum circuit generation; the qualitative thesis is sound, but the formal hardness claims are conditional and one proof is wrong as written. read the letter →

arxiv 2607.15313 v1 pith:QWPOBPQS submitted 2026-07-15 cs.LG

Junhao Song , Yu Zhou , William Knottenbelt , Yudong Cao This is my paper
classification cs.LG
keywords quantumprogramgenerationscalinghypothesisinversecircuitsynthesisverifier-centricsyntax-semanticsgapfunctionalcorrectnessfault-tolerantcompilation
verification ladder T0 review T1 audit T2 compute T3 formal

The pith

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The reading

The paper argues that the scaling hypothesis—more parameters yield emergent reasoning—fails for generic quantum circuit synthesis. Because physically valid circuits are exponentially rare among all syntactically parseable circuits, generating many candidates and then filtering out invalid ones has an expected cost that grows exponentially in qubit count, no matter how large the model is. Training on unverified quantum code teaches syntax without Hilbert-space semantics, so larger models can become more confidently wrong. The paper's proposed remedy is verifier-centric generation: build formal checks for structural and functional validity into the decoding process itself, via a masked policy that only permits actions that preserve correctness. If the argument holds, progress in quantum program generation depends on coupling models to verifiers and curated verified data, not on continued scaling.

What carries the argument

The central object is the two-layer validity decomposition into structural validity Mstruct and functional correctness Mfunc, combined in the coupled-exponential post-selection cost identity E[C] = Cstruct/(pstruct·pfunc|struct) + Cfunc/pfunc|struct, where Cfunc = Θ(2^n), Cstruct = Θ(nd), and pfunc|struct ≤ Cθ·e^{-γn}. The companion mechanism is the constrained policy πθ(a|s) ∝ πθ^(0)(a|s)·V(a,s), a validity mask over a Markov decision process over circuit states, which turns the verifier from an output gate into a generation invariant.

What would settle it

Benchmark a model on random n-qubit target unitaries (n up to about 20) that are provably absent from training data, and measure the conditional functional pass rate as n grows. The paper predicts exponential decay e^{-γn}; observing polynomial decay or a constant pass rate would refute the central claim. Also estimate the model's probability mass on the exact-solution region: if this mass grows exponentially with n even for random targets, the non-pathology assumption Cθ = O(poly(n)) fails.

Watch

Extended reading notes

Core claim

The paper's central claim is that post-hoc verification—generate candidates, then discard those that fail to compile or fail to implement the target unitary—is exponentially intractable for generic synthesis. Validity is split into two layers: structural validity (obeying hardware constraints) and functional correctness (implementing the target unitary within tolerance). The expected cost of obtaining one valid candidate follows a coupled-exponential identity: the structural verification cost is divided by the product of the structural and functional pass rates, while the functional verification cost, which is exponential in qubit count, is divided only by the functional pass rate. Because t

Load-bearing premise

The exponential-wall argument assumes the model has not memorized the specific target circuit; for textbook primitives like quantum Fourier transform or modular arithmetic that assumption fails, and in those cases scale plus data can still work.

Editorial extensions

If this is right

  • For generic synthesis, larger models trained on unverified code will not converge to physical validity; at best they produce more fluent but still invalid circuits, a predicted inverse-scaling regime.
  • Generate-then-verify pipelines have expected cost exponential in the logical qubit count (e^{(ln2 + γ)n} for NISQ, e^{(max(ln2,β)+γ)n} for FTQC), so post-hoc filtering is not a viable route to reliable quantum code.
  • Verification must be moved inside generation: masks, hierarchical module constraints, and property-preserving rewrites replace sampling-and-discarding.
  • Verified, trace-based training data becomes a primary design component, not a curation afterthought; the agent's target audience shifts from human reviewer to formal verifier.
  • The validity gap is independent of physical substrate: NISQ and FTQC both face the same exponential functional-sparsity barrier because the logical Hilbert-space dimension is what matters.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

  • A testable extension is to measure the conditional pass rate on a family of random n-qubit target unitaries not in the training set; the paper's bound predicts exponential decay with an exponent tracking depth and log(1/ε) or log n, checkable by state-vector verification up to about 20 qubits.
  • The coupled-exponential argument generalizes to other engineering-science domains where valid outputs are exponentially rare and functional verification is exponential—protein design, theorem proving, and chip layout are analogous, and the same verifier-in-the-loop recipe should apply there.
  • The sharpest consequence the paper leaves implicit is that scale investments in quantum-code language models only pay off when paired with verification-aware architectures; absent that, continued scaling is predicted to be actively harmful rather than merely flat.
  • The framework implies benchmark design should shift from token-level accuracy to verified-synthesis pass rates at a fixed verification budget, a shift that would make the paper's central claim directly measurable across different model families.
Share X Bluesky LinkedIn Reddit HN

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Referee Report

3 major / 4 minor

Summary. This position paper argues that standard probabilistic scaling of large language models is the wrong paradigm for generic quantum circuit synthesis. It identifies a syntax-semantics gap, claims that training on unverified quantum code biases models away from physical validity, and argues that post-hoc verification is intractable because valid circuits are exponentially sparse. The paper formalizes a two-layer validity notion (structural/functional) and derives a coupled exponential cost for generate-then-filter pipelines (Eq. (2)), then proposes verifier-centric agents that enforce validity during generation via hierarchical constraints, validity masks, and trace-based data. Supporting appendices contain formal sparsity propositions for NISQ and FTQC, a structural-hardness claim, and a worked Cuccaro adder example.

Significance. The paper addresses an important and timely question: whether LLM scaling alone can produce correct quantum programs. The qualitative thesis is plausible and aligns with recent evidence in constrained domains such as formal mathematics and planning. The proposal to make verification a generation invariant rather than a post-filter is a useful contribution for quantum software and AI4Science. The paper is also commendably explicit about several assumptions (footnote 1, Appendix C.2 remarks), and the hierarchical multi-level abstraction framework in Section 4 is concrete. However, the central impossibility claim is currently gated by assumptions that are not established, and the structural-hardness argument contains a technical error. With those repaired or explicitly scoped, the paper could be a strong position statement. No machine-checked proofs or reproducible artifacts are provided; the empirical anchor is a self-cited workshop paper.

major comments (3)
  1. [Appendix B.3, Proposition B.1] The amplification step is incorrect. The proof supposes P(x∈M_struct^FTQC)≥2^{-cn} and claims that 'poly(n) independent repetitions' followed by poly(n)-time verification yields constant success probability, contradicting NP⊈BPP. In fact, poly(n) repetitions of a 2^{-cn} event succeed with probability poly(n)·2^{-cn}=e^{-Ω(n)}, still exponentially small; constant success requires 2^{cn} repetitions. The earlier ETH argument is also wrong: 2^{cn}=2^{Θ(m)} expected trials is not o(2^m) for any constant c>0, so no ETH contradiction follows. The reduction step asserting a poly-time bijection between structurally valid FTQC descriptions and satisfying assignments is asserted without proof. Because Eq. (4) and POSITION2 rely on p_struct^FTQC≤e^{-βn}, this is a load-bearing gap. Please correct the statement (e.g., conditional on a sampler achieving 1/poly(n) mass) or provide a valid reduction/a
  2. [Section 3.2 / Eq. (1) / footnote 1] The functional-sparsity bound is informative only when Cθ=O(poly(n)), which the paper takes as holding 'by definition of generalization for out-of-distribution targets.' This is an assumption, not a demonstrated property, and it is the gate for the paper's central negative claim. The paper concedes that for in-distribution algorithmic primitives Cθ can be large enough to cancel e^{-γn}; nothing in the argument rules out a scaled model moving currently-OOD targets into its effective support and giving Cθ=e^{Ω(n)}. The cited empirical evidence (Song et al., 2026) is a self-cited workshop paper reporting a 79% ceiling, not a scaling-law study showing the ceiling persists with model size. POSITION2 should therefore be stated as a conditional conjecture under the non-memorization condition Cθ=O(poly(n)), or the paper should supply a scaling-dependent bound on Cθ and a concrete falsifiable exp
  3. [Appendix C.2.2, Eq. (19)] The proof of Proposition C.2 does not establish the claimed upper bound on the number of length-G programs that ε-approximate U_target. The bound poly(n,log(1/ε))·|I|^{G-G*} assumes every such program is a no-op/rewrite extension of a canonical length-G* program; an arbitrary length-G program approximating U_target need not contain such a subprogram, and there can be exponentially many distinct programs approximating the identity that could be appended. Without a rigorous counting argument, the numerator bound and hence γ_FTQC=Ω(d log n) are not proven. This is a second load-bearing gap in the formal core of POSITION2. Please provide a complete argument or label Proposition C.2 as a heuristic scaling estimate.
minor comments (4)
  1. [Appendix A] The Pauli-stabilizer evolution appears incorrect: starting from ZII and applying CNOT(y,x), CNOT(x,temp), CNOT(y,x) gives ZZZ, not ZIZ as stated. The qualitative conclusion that temp is entangled still holds, but the displayed calculation should be corrected.
  2. [Abstract / Section 3.3] The phrase 'post-hoc filtering is mathematically intractable' is stronger than the formal results, which are conditional on Cθ=O(poly(n)) and on the corrected structural-hardness argument. Consider qualifying the abstract and POSITION2 accordingly.
  3. [Appendix C.2.2] The assumption that U_target admits a polynomial-length canonical description is stated only in the setup, not in Proposition C.2. State it in the proposition statement.
  4. [Section 2.2] The key empirical ceiling (79%) is cited to the authors' own workshop paper (Song et al., 2026). Please provide more experimental detail (model sizes, tasks, error bars) or cite independent evidence, since this is a primary motivation for the inverse-scaling hypothesis.

Circularity Check

1 steps flagged · score 6.0 of 10

Central impossibility claim is gated by a definitionally imposed bound on Cθ; without it, Eq. (1) does not rule out scaling-induced concentration.

  1. self definitional [Section 3.2, footnote 1; Appendix C.2, Remark (i); POSITION2 (Section 3.3)]
    "The bound is pointwise in Utarget but informative only when Cθ = O(poly(n)) , which holds by definition of generalization for out-of-distribution targets. For in-distribution algorithmic primitives (textbook QFT, Grover, modular arithmetic), Cθ can be large enough to cancel e−γn entirely..."

    Equation (1) is the only mechanism supplied for POSITION2's claim that 'probabilistic scaling cannot close the coupled exponential gap.' Its exponential decay is multiplied by Cθ; a larger model placing more mass near an out-of-distribution target makes Cθ large and can cancel e^{−γn}. The paper does not prove scaling cannot do this; it asserts Cθ=O(poly(n)) 'by definition of generalization for out-of-distribution targets.' The no-scaling conclusion is therefore equivalent to assuming the learned distribution stays near-uniform at all scales, which is exactly the quantity scaling is supposed to affect. The impossibility result is a conditional conjecture, not a demonstrated barrier.

full rationale

The derivation of the coupled exponential cost (Eqs. (1)-(4)) has substantial independent content: Haar-measure sparsity, the Θ(2^n) verification cost, and the structural-hardness reduction are real ingredients. If Cθ≤poly(n) were independently established for all scales, POSITION2 would follow. But the paper's own footnote restricts the bound to that assumption and labels it 'by definition of generalization.' That is the central circular step: the no-scaling conclusion is equivalent to assuming the model cannot concentrate probability mass on out-of-distribution targets, which is precisely what scaling is hypothesized to improve. The self-citation to Song et al. (2026) for the 79% ceiling is ancillary support for the inverse-scaling hypothesis and does not itself carry the derivation, so I do not count it as load-bearing circularity. Appendix B.3 contains a separate, non-circular flaw—poly(n) repetitions cannot amplify a 2^{−cn} success probability to constant—which affects Proposition B.1 but is a correctness issue rather than a circularity. Overall: partial circularity in the central claim, while the constructive verifier-centric proposal remains independent content.

Assumptions & free parameters 3 free parameters · 6 assumptions · 0 invented entities

The paper introduces no new physical entities. Its analytical payload consists of asymptotic bounds and an agent architecture. The main free parameters are the non-pathology constants Cθ and β, which are assumed rather than measured. The domain assumptions are standard in the quantum-verification literature, while the regularity assumptions (R1)-(R2) and the Cθ definition are introduced specifically to make the proofs go through.

free parameters (3)
  • Cθ = assumed O(poly(n)) for out-of-distribution targets
    Measures how far πθ deviates from uniform; the exponential bounds only apply when Cθ is small, and the paper defines this to hold for targets the model has not memorized (Appendix C.2, Eq. 10 and Remark (i)).
  • β = positive constant in e^{-βn} for FTQC structural sparsity
    Implicit constant in Proposition B.1's Ω(n) exponent; never quantified, but Equation (4) uses it as if known.
  • ϵ = empirically driven below 1/G in NISQ
    Per-gate structural violation probability in the NISQ pass rate (1−ϵ)^G; the paper states fine-tuning reaches this regime citing other systems, but no measurement is provided.
assumptions (6)
  • ad hoc to paper The map from ansatz parameters to unitaries is generically full-rank with j0=Ω(1) and bounded fiber multiplicity (R1)-(R2).
    Required by Proposition C.1; barren-plateau ansatzes are explicitly excluded (Appendix C.2.1, Remark (v)).
  • domain assumption Useful post-classical quantum algorithms have volume-law entanglement, forcing tensor-network bond dimension e^{Ω(n)}.
    Used to rule out MPS-based verification for generic synthesis (Section 3.2, Appendix C.1).
  • standard math Verification of arbitrary n-qubit circuits costs Θ(2^n) classically; non-identity testing is QMA-complete.
    Basis for Cfunc in Section 3.2 and Background 2, citing Watrous 2009.
  • ad hoc to paper Classical 3-SAT can be reduced to FTQC structural feasibility in polynomial time with a bijection to satisfying assignments.
    Needed for Proposition B.1; only sketched via qubit-allocation NP-hardness (Appendix B.3).
  • ad hoc to paper Cθ=O(poly(n)) defines out-of-distribution generalization.
    Gates the applicability of the functional sparsity bound to non-memorized targets (Appendix C.2, Remark (i)).
  • domain assumption Training samples from unverified public quantum corpora create structural bias against physical validity.
    Central premise of the 'data poisoning' argument (Section 2.2).

how reviews work

0 comments
Cite this review

Pith. "Pith review of Position: Quantum Program Generation Must Prioritize Validity Over Probabilistic Scaling." pith.science (2026). https://pith.science/paper/QWPOBPQS

@misc{pith2026260715313,
  author       = {Pith},
  title        = {Pith review of: Position: Quantum Program Generation Must Prioritize Validity Over Probabilistic Scaling},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/QWPOBPQS}},
  note         = {Machine review of arXiv:2607.15313}
}
read the original abstract

The scaling hypothesis assumes that increasing model parameters yields emergent reasoning capabilities. This position paper argues that applying this probabilistic paradigm to generic quantum circuit synthesis is a directional error. Unlike natural languages, quantum circuits require strict adherence to mathematical constraints that manifest a significant syntax-semantics gap. Training on unverified quantum programs means that models learn syntax but fail to capture the physical semantics of the Hilbert space. Since the valid subset of circuit designs decays exponentially with the number of qubits, post-hoc filtering is mathematically intractable. We propose a pivot from human-centric copilots to verifier-centric agents. We integrate hierarchical constraints, topological masks, and symbolic proxies directly into generation. Our analysis suggests that scale alone cannot bridge the validity gap. Verification-aware architectures offer a viable path for modular quantum program generation. These considerations point toward generation methods that encode task-specific rules of quantum information, rather than relying on imitation alone.

Figures

Figures reproduced from arXiv: 2607.15313 by the authors.

Figure 1
Figure 1. Comparison of the traditional method and reliable LLM with closed loop verification. When data verification is integrated into the training loop, only validated designs from the constrained validation space are used, ensuring higher correct rate outputs. Recent systems like IBM’s Qiskit Code Assistant (Dupuis et al., 2024) and DeepMind’s AlphaTensor-Quantum (Ruiz et al., 2025) show promise. However, a critical disse… view at source ↗
Figure 2
Figure 2. Toy example contrasting classical and quantum pro￾gramming: syntactic fluency is much less of a predictor of seman￾tic correctness in quantum programming than in classical program￾ming. ing pipeline (Dupuis et al., 2024), code from official Qiskit GitHub organizations is oversampled by 10.3× relative to general scraped repositories, a ratio chosen empirically for benchmark performance. We take such filtering as a st… view at source ↗
Figure 3
Figure 3. Two-stage post hoc filtering for quantum synthesis. routes must not cross in spacetime, factory throughput must meet T-gate demand at every timestep, and distance budgets must sum to the target logical error rate (Litinski, 2019; Bev￾erland et al., 2022). These constraints are not separable, and the corresponding feasibility problems inherit NP-hardness from the qubit-mapping subroutine (Siraichi et al., 2018; Botea… view at source ↗
Figures from the paper (2 more)
Figure 4
Figure 4. Figure 4: Control flow of the verification protocol. This can be considered as an expanded view of [PITH_FULL_IMAGE:figures/full_fig_p007_4.png]
Figure 5
Figure 5. Figure 5: Illustration of the different levels of abstraction using the Cuccaro Adder (Cuccaro et al., 2004) as an example. (a) Level 1 in the hierarchy describes the abstract circuit design based on modules MAJ and UMA. Here a concrete example with n = 2 is shown. (b) Level 2 e…

Discussion (0). Sign in to comment.

Reference graph

Works this paper leans on

93 extracted references · 21 linked inside Pith

  1. [1]

    arXiv preprint arXiv:2108.07732 , year=

    Program Synthesis with Large Language Models , author=. arXiv preprint arXiv:2108.07732 , year=

    arXiv

  2. [2]

    arXiv preprint arXiv:2107.03374 , year=

    Evaluating Large Language Models Trained on Code , author=. arXiv preprint arXiv:2107.03374 , year=

    arXiv

  3. [3]

    Qiskit code assistant: Training

    Dupuis, Nicolas and Buratti, Luca and Vishwakarma, Sanjay and Forrat, Aitana Viudes and Kremer, David and Faro, Ismael and Puri, Ruchir and Cruz-Benito, Juan , booktitle=. Qiskit code assistant: Training. 2024 , organization=

    2024

  4. [4]

    Quantum circuit optimization with

    Ruiz, Francisco JR and Laakkonen, Tuomas and Bausch, Johannes and Balog, Matej and Barekatain, Mohammadamin and Heras, Francisco JH and Novikov, Alexander and Fitzpatrick, Nathan and Romera-Paredes, Bernardino and van de Wetering, John and others , journal=. Quantum circuit optimization with. 2025 , publisher=

    2025

  5. [5]

    Nature , volume =

    Mathematical discoveries from program search with large language models , author =. Nature , volume =. 2024 , doi =

    2024

  6. [6]

    2024 , doi=

    Vishwakarma, Sanjay and Harkins, Francis and Golecha, Siddharth and Bajpe, Vishal Sharathchandra and Dupuis, Nicolas and Buratti, Luca and Kremer, David and Faro, Ismael and Puri, Ruchir and Cruz-Benito, Juan , journal=. 2024 , doi=

    2024

  7. [7]

    arXiv preprint arXiv:2001.08361 , year=

    Scaling laws for neural language models , author=. arXiv preprint arXiv:2001.08361 , year=

    arXiv 2001

  8. [8]

    34th European Conference on Object-Oriented Programming (ECOOP 2020) , series =

    Don't Panic! Better, Fewer, Syntax Errors for LR Parsers , author =. 34th European Conference on Object-Oriented Programming (ECOOP 2020) , series =. 2020 , publisher =

    2020

Show all 93 references
  1. [9]

    Puterman , title =

    Martin L. Puterman , title =. 1994 , series =. doi:10.1002/9780470316887 , isbn =

    1994 doi

  2. [10]

    Sutton and Andrew G

    Richard S. Sutton and Andrew G. Barto , title =. 2018 , doi =

    2018

  3. [11]

    arXiv preprint arXiv:2405.09178 , year=

    Testing and Debugging Quantum Programs: The Road to 2030 , author=. arXiv preprint arXiv:2405.09178 , year=

    arXiv

  4. [12]

    and Draper, Thomas G

    Cuccaro, Steven A. and Draper, Thomas G. and Kutin, Samuel A. and Moulton, David Petrie , title =. doi:10.48550/arxiv.quant-ph/0410184 , publisher =

    doi arXiv

  5. [13]

    Transpiler (latest version) |. n.d. , note =

  6. [14]

    Kambhampati, Subbarao and Valmeekam, Karthik and Guan, Lin and Verma, Mudit and Stechly, Kaya and Bhambri, Siddhant and Saldyt, Lucas and Murthy, Anil , journal=

  7. [15]

    Advances in neural information processing systems , volume=

    Attention is all you need , author=. Advances in neural information processing systems , volume=

  8. [16]

    , title =

    Bennett, Charles H. , title =. IBM Journal of Research and Development , volume =. 1973 , month = nov, doi =

    1973

  9. [17]

    An Evaluation Benchmark for Autoformalization in

    Gulati, Aryan and Ladsaria, Devanshu and Mishra, Shubhra and Sidhu, Jasdeep and Miranda, Brando , journal=. An Evaluation Benchmark for Autoformalization in

  10. [18]

    APOLLO: Automated

    Ospanov, Azim and Farnia, Farzan and Yousefzadeh, Roozbeh , journal=. APOLLO: Automated

  11. [19]

    arXiv preprint arXiv:2309.07062 , year=

    Large Language Models for Compiler Optimization , author=. arXiv preprint arXiv:2309.07062 , year=

    arXiv

  12. [20]

    Transferable Presynthesis PPA Estimation for

    Fang, Wenji and Lu, Yao and Liu, Shang and Zhang, Qijun and Xu, Ceyu and Wills, Lisa Wu and Zhang, Hongce and Xie, Zhiyao , journal=. Transferable Presynthesis PPA Estimation for. 2024 , publisher=

    2024

  13. [21]

    Abstraction-of-Thought: Intermediate Representations for

    DeLorenzo, Matthew and Tieu, Kevin and Jana, Prithwish and Jha, Piyush and Kalathil, Dileep and Ganesh, Vijay and Rajendran, Jeyavijayan , journal=. Abstraction-of-Thought: Intermediate Representations for

  14. [22]

    Veriopt:

    Tasnia, Kimia and Garcia, Alexander and Farheen, Tasnuva and Rahman, Sazadur , booktitle=. Veriopt:. 2025 , organization=

    2025

  15. [23]

    TMLR , year=

    Emergent abilities of large language models , author=. TMLR , year=

  16. [24]

    TMLR , year=

    Inverse scaling: When bigger isn't better , author=. TMLR , year=

  17. [25]

    Nature , volume=

    Evidence for the utility of quantum computing before fault tolerance , author=. Nature , volume=. 2023 , publisher=

    2023

  18. [26]

    Competition-level code generation with

    Li, Yujia and Choi, David and Chung, Junyoung and Kushman, Nate and Schrittwieser, Julian and Leblond, R. Competition-level code generation with. Science , volume=. 2022 , publisher=

    2022

  19. [27]

    arXiv preprint arXiv:2005.06787 , year=

    Classical Simulation of Quantum Supremacy Circuits , author=. arXiv preprint arXiv:2005.06787 , year=

    2005 arXiv

  20. [28]

    Quantum , volume=

    How many qubits are needed for quantum computational supremacy? , author=. Quantum , volume=. 2020 , doi=

    2020

  21. [29]

    Nature , volume=

    Quantum supremacy using a programmable superconducting processor , author=. Nature , volume=. 2019 , doi=

    2019

  22. [30]

    Quantum , volume=

    Quantum computing in the NISQ era and beyond , author=. Quantum , volume=. 2018 , publisher=

    2018

  23. [31]

    2010 , publisher=

    Quantum Computation and Quantum Information: 10th Anniversary Edition , author=. 2010 , publisher=

    2010

  24. [32]

    Advances in Neural Information Processing Systems , volume=

    Training compute-optimal large language models , author=. Advances in Neural Information Processing Systems , volume=. 2022 , doi=

    2022

  25. [33]

    Nature , volume=

    A single quantum cannot be cloned , author=. Nature , volume=. 1982 , doi=

    1982

  26. [34]

    Proceedings of the 2019 ACM SIGPLAN International Symposium on New Ideas, New Paradigms, and Reflections on Programming and Software , pages=

    The adverse effects of code duplication in machine learning models of code , author=. Proceedings of the 2019 ACM SIGPLAN International Symposium on New Ideas, New Paradigms, and Reflections on Programming and Software , pages=. 2019 , doi=

    2019

  27. [35]

    Proceedings of the 44th International Conference on Software Engineering , pages=

    On the importance of building high-quality training datasets for neural code search , author=. Proceedings of the 44th International Conference on Software Engineering , pages=. 2022 , doi=

    2022

  28. [36]

    ACM Computing Surveys , volume=

    Survey of hallucination in natural language generation , author=. ACM Computing Surveys , volume=. 2023 , publisher=

    2023

  29. [37]

    Climbing towards

    Bender, Emily M and Koller, Alexander , booktitle=. Climbing towards. 2020 , doi=

    2020

  30. [38]

    Encyclopedia of Complexity and Systems Science , pages=

    Quantum computational complexity , author=. Encyclopedia of Complexity and Systems Science , pages=. 2009 , publisher=

    2009

  31. [39]

    Introduction to

    Mele, Antonio Anna , journal=. Introduction to. 2024 , doi=

    2024

  32. [40]

    Physical Review Letters , volume=

    Entropy scaling and simulability by matrix product states , author=. Physical Review Letters , volume=. 2008 , doi=

    2008

  33. [41]

    arXiv preprint arXiv:1607.05256 , year=

    The complexity of quantum states and transformations: from quantum money to black holes , author=. arXiv preprint arXiv:1607.05256 , year=

    arXiv

  34. [42]

    Nature , volume=

    Quantum computing with realistically noisy devices , author=. Nature , volume=. 2005 , doi=

    2005

  35. [43]

    Physical Review Letters , volume=

    Efficient classical simulation of slightly entangled quantum computations , author=. Physical Review Letters , volume=. 2003 , doi=

    2003

  36. [44]

    arXiv preprint arXiv:1411.4028 , year=

    A quantum approximate optimization algorithm , author=. arXiv preprint arXiv:1411.4028 , year=

    arXiv

  37. [45]

    arXiv preprint arXiv:2310.13296 , year=

    Trotterization in Quantum Theory , author=. arXiv preprint arXiv:2310.13296 , year=

    arXiv

  38. [46]

    Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing , pages=

    Constrained Language Models Yield Few-Shot Semantic Parsers , author=. Proceedings of the 2021 Conference on Empirical Methods in Natural Language Processing , pages=. 2021 , doi=

    2021

  39. [47]

    arXiv preprint arXiv:2305.20050 , year=

    Let's verify step by step , author=. arXiv preprint arXiv:2305.20050 , year=

    arXiv

  40. [48]

    New Journal of Physics , volume=

    Interacting quantum observables: categorical algebra and diagrammatics , author=. New Journal of Physics , volume=. 2011 , doi=

    2011

  41. [49]

    arXiv preprint quant-ph/9807006 , year=

    The Heisenberg representation of quantum computers , author=. arXiv preprint quant-ph/9807006 , year=

    arXiv

  42. [50]

    Electronic Proceedings in Theoretical Computer Science , volume=

    PyZX: Large scale automated diagrammatic reasoning , author=. Electronic Proceedings in Theoretical Computer Science , volume=. 2020 , doi=

    2020

  43. [51]

    Nature , volume=

    Larger and more instructable language models become less reliable , author=. Nature , volume=. 2024 , doi=

    2024

  44. [52]

    Advances in Neural Information Processing Systems , volume=

    Faith and fate: Limits of transformers on compositionality , author=. Advances in Neural Information Processing Systems , volume=. 2023 , doi=

    2023

  45. [53]

    Proceedings of the ACM on Programming Languages , volume=

    A verified optimizer for quantum circuits , author=. Proceedings of the ACM on Programming Languages , volume=. 2021 , doi=

    2021

  46. [54]

    Proceedings of the 28th Annual ACM Symposium on Theory of Computing , pages=

    A fast quantum mechanical algorithm for database search , author=. Proceedings of the 28th Annual ACM Symposium on Theory of Computing , pages=. 1996 , doi=

    1996

  47. [55]

    arXiv preprint arXiv:1803.04402 , year=

    Quantum supremacy and the complexity of random circuit sampling , author=. arXiv preprint arXiv:1803.04402 , year=

    arXiv

  48. [56]

    arXiv preprint arXiv:2502.20246 , year=

    Beyond Natural Language Perplexity: Detecting Dead Code Poisoning in Code Generation Datasets , author=. arXiv preprint arXiv:2502.20246 , year=

    arXiv

  49. [57]

    Quantum computing with

    Javadi-Abhari, Ali and Treinish, Matthew and Krsulich, Kevin and Wood, Christopher J and Lishman, Jake and Gacon, Julien and Martiel, Simon and Nation, Paul D and Bishop, Lev S and Cross, Andrew W and others , journal=. Quantum computing with

  50. [58]

    A Closer Look at Invalid Action Masking in Policy Gradient Algorithms , volume=

    Huang, Shengyi and Ontañón, Santiago , year=. A Closer Look at Invalid Action Masking in Policy Gradient Algorithms , volume=. doi:10.32473/flairs.v35i.130584 , journal=

    doi

  51. [59]

    Applied Sciences , year =

    Yueqi Hou and Xiaolong Liang and Jiaqiang Zhang and Qisong Yang and Aiwu Yang and Ning Wang , title =. Applied Sciences , year =

  52. [60]

    Computer Science Review , volume=

    A survey on quantum computing technology , author=. Computer Science Review , volume=. 2019 , doi=

    2019

  53. [61]

    Findings of the Association for Computational Linguistics: ACL 2023 , pages=

    The larger they are, the harder they fail: Language models do not recognize identifier swaps in python , author=. Findings of the Association for Computational Linguistics: ACL 2023 , pages=

    2023

  54. [62]

    Quantum , volume=

    A game of surface codes: Large-scale quantum computing with lattice surgery , author=. Quantum , volume=. 2019 , publisher=

    2019

  55. [63]

    arXiv preprint arXiv:1808.06709 , year=

    Low overhead quantum computation using lattice surgery , author=. arXiv preprint arXiv:1808.06709 , year=

    arXiv

  56. [64]

    , title =

    Herr, Daniel and Nori, Franco and Devitt, Simon J. , title =. npj Quantum Information , volume =. 2017 , doi =

    2017

  57. [65]

    arXiv preprint arXiv:2512.06484 , year=

    Scheduling Lattice Surgery with Magic State Cultivation , author=. arXiv preprint arXiv:2512.06484 , year=

    arXiv

  58. [66]

    , title =

    Ding, Yongshan and Holmes, Adam and Javadi-Abhari, Ali and Franklin, Diana and Martonosi, Margaret and Chong, Frederic T. , title =. 51st Annual IEEE/ACM International Symposium on Microarchitecture (MICRO) , pages =. 2018 , doi =

    2018

  59. [67]

    arXiv preprint arXiv:2401.15829 , year =

    Hamada, Kou and Suzuki, Yasunari and Tokunaga, Yuuki , title =. arXiv preprint arXiv:2401.15829 , year =

    arXiv

  60. [68]

    Proceedings of the 24th International Conference on Architectural Support for Programming Languages and Operating Systems (ASPLOS '19) , pages =

    Li, Gushu and Ding, Yufei and Xie, Yuan , title =. Proceedings of the 24th International Conference on Architectural Support for Programming Languages and Operating Systems (ASPLOS '19) , pages =. 2019 , doi =

    2019

  61. [69]

    and others , title =

    Varbanov, Boris M. and others , title =. arXiv preprint arXiv:2502.03908 , year =

    arXiv

  62. [70]

    Non-Identity-Check

    "Non-Identity-Check" is QMA-Complete , author=. International Journal of Quantum Information , volume=. 2005 , doi=

    2005

  63. [71]

    arXiv preprint arXiv:0906.5416 , year=

    Non-Identity Check Remains QMA-Complete for Short Circuits , author=. arXiv preprint arXiv:0906.5416 , year=

    arXiv

  64. [72]

    arXiv preprint arXiv:2211.07629 , year=

    Assessing requirements to scale to practical quantum advantage , author=. arXiv preprint arXiv:2211.07629 , year=

    arXiv

  65. [73]

    Proceedings of the 2018 international symposium on code generation and optimization , pages=

    Qubit allocation , author=. Proceedings of the 2018 international symposium on code generation and optimization , pages=

    2018

  66. [74]

    Proceedings of the International Symposium on Combinatorial Search , volume=

    On the complexity of quantum circuit compilation , author=. Proceedings of the International Symposium on Combinatorial Search , volume=

  67. [75]

    2024 ACM/IEEE 51st Annual International Symposium on Computer Architecture (ISCA) , pages=

    A sat scalpel for lattice surgery: representation and synthesis of subroutines for surface-code fault-tolerant quantum computing , author=. 2024 ACM/IEEE 51st Annual International Symposium on Computer Architecture (ISCA) , pages=. 2024 , organization=

    2024

  68. [76]

    Algorithms and Data Structures Symposium , pages=

    Algorithmic theory of qubit routing , author=. Algorithms and Data Structures Symposium , pages=. 2023 , organization=

    2023

  69. [77]

    arXiv preprint arXiv:2405.17688 , year=

    Multi-qubit lattice surgery scheduling , author=. arXiv preprint arXiv:2405.17688 , year=

    arXiv

  70. [78]

    arXiv preprint LAUR-95-2225 , year =

    Knill, Emanuel , title =. arXiv preprint LAUR-95-2225 , year =

  71. [79]

    1969 , series =

    Federer, Herbert , title =. 1969 , series =

    1969

  72. [80]

    and Nielsen, Michael A

    Dawson, Christopher M. and Nielsen, Michael A. , title =. Quantum Information & Computation , volume =. 2006 , eprint =

    2006

  73. [81]

    Transactions on Machine Learning Research , year=

    Inverse Scaling in Test-Time Compute , author=. Transactions on Machine Learning Research , year=

  74. [82]

    In 40th Proceedings of the Association for the Advancement of Artificial Intelligence AAAI 2026 AI4Research Workshop , year=

    Data Verification is the Future of Quantum Computing Copilots , author=. In 40th Proceedings of the Association for the Advancement of Artificial Intelligence AAAI 2026 AI4Research Workshop , year=

    2026

  75. [83]

    Song, Junhao and Yuan, Yingfang and Pang, Wei , booktitle=

  76. [84]

    and Boixo, Sergio and Smelyanskiy, Vadim N

    McClean, Jarrod R. and Boixo, Sergio and Smelyanskiy, Vadim N. and Babbush, Ryan and Neven, Hartmut , title =. Nature Communications , volume =. 2018 , doi =

    2018

  77. [85]

    and Sone, Akira and Volkoff, Tyler and Cincio, Lukasz and Coles, Patrick J

    Cerezo, M. and Sone, Akira and Volkoff, Tyler and Cincio, Lukasz and Coles, Patrick J. , title =. Nature Communications , volume =. 2021 , doi =

    2021

  78. [86]

    and Polikarpova, Nadia , title =

    Barke, Shraddha and James, Michael B. and Polikarpova, Nadia , title =. Proceedings of the ACM on Programming Languages , volume =. 2023 , doi =

    2023

  79. [87]

    , title =

    Vaithilingam, Priyan and Zhang, Tianyi and Glassman, Elena L. , title =. Extended Abstracts of the 2022 CHI Conference on Human Factors in Computing Systems , series =. 2022 , pages =

    2022

  80. [88]

    and Yang, Chenyang and Myers, Brad A

    Liang, Jenny T. and Yang, Chenyang and Myers, Brad A. , title =. Proceedings of the 46th IEEE/ACM International Conference on Software Engineering (ICSE) , year =

  81. [89]

    2021 , publisher=

    Scholak, Torsten and Schucher, Nathan and Bahdanau, Dzmitry , booktitle=. 2021 , publisher=

    2021

  82. [90]

    2018 IEEE 59th Annual Symposium on Foundations of Computer Science (FOCS) , pages=

    Classical verification of quantum computations , author=. 2018 IEEE 59th Annual Symposium on Foundations of Computer Science (FOCS) , pages=. 2018 , organization=

    2018

  83. [91]

    Physical Review A , volume=

    Unconditionally verifiable blind quantum computation , author=. Physical Review A , volume=. 2017 , doi=

    2017

  84. [92]

    Nature , volume=

    Classical command of quantum systems , author=. Nature , volume=. 2013 , doi=

    2013

  85. [93]

    Annual International Cryptology Conference , pages=

    Succinct classical verification of quantum computation , author=. Annual International Cryptology Conference , pages=. 2022 , organization=

    2022

Pith tools

Reviewed August 2, 2026 · model on record in the stance chip above.