{"id":"69070cc1-3bec-4d16-b276-31c49ed1ae12","arxiv_id":"2508.05523","paper_version":3,"verdict":"UNVERDICTED","confidence":"LOW","novelty_score":7.0,"correctness_risk":"unknown","formal_verification":"none","parameter_count":2,"one_line_summary":"Introduces logical accreditation, a randomized compilation based protocol that turns logical circuit noise into stochastic Pauli noise, giving scalable certification and bounds for fault-tolerant quantum computations.","lead":"Logical accreditation is a new framework for checking whether computations performed on error-corrected quantum hardware are correct, avoiding the need to classically simulate the computation. It matters because as fault-tolerant quantum computers grow, classical verification becomes intractable, and this is a candidate scalable audit method for them.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"No significant objection identified","rationale":"The reader's verdict is UNVERDICTED, low confidence, based solely on the abstract because the full text is unreadable. The stress-test pass cannot do better: the manuscript body is corrupted and no technical argument can be verified. The reader's weakest_assumption—that the twirling procedure maps arbitrary logical noise to stochastic Pauli noise for non-transversal gates—is indeed the most fragile premise, but it cannot be confirmed or refuted from the available material. Therefore, no load-bearing concern can be identified, and the appropriate action is to leave the verdict unchanged. The partial agreement reflects that we acknowledge this premise is the critical point to check, even though we do not assert that it fails. The concrete test proposed would provide the necessary evidence once a readable version is available.","tokens_in":21126,"tokens_out":8427,"duration_ms":93755,"concrete_test":"Obtain a readable version of the paper and independently apply the randomized compilation prescription to a concrete non-transversal logical gate (e.g., logical Toffoli in a small code) under a realistic physical noise model. Compute the averaged logical channel and check whether it is exactly a stochastic Pauli channel; if non-Pauli terms persist, the framework's foundational equivalence fails.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The provided full text is corrupted (mojibake; embedded header cites arXiv:2508.05526v2 [cs.CV], not the target quant-ph paper), so the central derivations cannot be audited. The abstract's claim that randomized compilation converts arbitrary logical circuit noise into stochastic Pauli noise, including non-transversal gates beyond T, is plausible but unverified. A specific technical concern would require inspecting the twirling construction and the statistical bound. Without a readable derivation, no load-bearing objection can be raised; the honest status is 'unverified,' consistent with the reader's UNVERDICTED verdict.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript proposes \"logical accreditation,\" a framework for certifying the correctness of fault-tolerant computations on logical qubits. The abstract claims that a novel randomized compilation scheme converts arbitrary logical circuit noise into stochastic Pauli noise, including a method for twirling non-transversal logical gates beyond the standard T gate, resolving an open problem posed by Piveteau et al. (PRL 127, 200505 (2021)). From this, the framework is said to upper-bound the infidelity of the logical output state, enable scalable certification without classical simulation, extend entropy benchmarking to fault-tolerant regimes, and numerically demonstrate crossover points where encoded computations outperform physical computations. The provided full text is unreadable (mojibake), so none of the technical derivations, proofs, or numerical details can be inspected.","tokens_in":21098,"tokens_out":4260,"duration_ms":42502,"significance":"If the claims are correct, the work would be significant: it would provide a scalable certification method for fault-tolerant computations under general noise models, without needing classical simulation, and would resolve a recognized open problem about twirling non-transversal gates. However, because the manuscript's technical content is inaccessible in the submitted form, the contribution currently exists only at the level of the abstract. No machine-checked proofs, reproducible code, or inspectable data are available to substantiate the claims.","major_comments":[{"comment":"The central claim—that the randomized compilation scheme converts arbitrary logical circuit noise into stochastic Pauli noise, including for non-transversal gates—is asserted in the abstract without any readable derivation. The full text is corrupted mojibake and includes a header citing arXiv:2508.05526v2 [cs.CV], not the target quant-ph paper. The twirling theorem and the resulting infidelity upper bound are load-bearing; without their proofs, the certification guarantee is unsupported. A readable manuscript with complete theorem statements and proofs is required.","section":"Full Text; Abstract"},{"comment":"The abstract reports numerical demonstrations of scalable certification and a crossover point, but no legible figures, tables, noise-model definitions, or code/data appear. Concretely, the physical error rates, non-Pauli noise components, number of compiled runs and shots, and statistical confidence intervals are unspecified. The numerical claims cannot be assessed or reproduced as presented.","section":"Abstract (numerical simulations)"},{"comment":"The abstract does not explain how Pauli error rates extracted from test runs certify a different target run. A load-bearing step is proving that the compiled circuit's effective noise is stochastic Pauli for every gate type, including non-transversal gates, and that finite-sample estimates provide a rigorous confidence bound. The manuscript must state the test/target independence conditions and any assumptions about temporally correlated or coherent noise; otherwise the claimed infidelity bound and scalability do not follow.","section":"Abstract (statistical control)"}],"minor_comments":[{"comment":"The embedded header 'arXiv:2508.05526v2 [cs.CV] 29 Dec 2025' does not match the target arXiv:2508.05523 (quant-ph); this indicates the provided text is corrupted and needs replacement.","section":"Full Text"},{"comment":"The term 'crossover point' is not defined in the abstract; a precise definition would help readers interpret the numerical claim about when encoded computations outperform physical computations.","section":"Abstract"},{"comment":"The citation to Piveteau et al., PRL 127, 200505 (2021), appears in the abstract, but the reference list cannot be read in the provided text, so the context of the 'open problem' is not verifiable.","section":"References"}],"recommendation":"uncertain","confidential_remarks":"I could not audit the manuscript because the full text is unreadable. This recommendation reflects lack of evidence rather than a detected technical error. If a readable version is provided, I would review the central twirling construction and the statistical infidelity bound carefully. The editor may wish to treat this as a request for a clean manuscript before normal peer review can proceed."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"You should know before reading this: the abstract promises a framework that converts arbitrary logical circuit noise into stochastic Pauli noise, twirls non-transversal gates beyond T (resolving an open problem from Piveteau et al. 2021), and upper-bounds output infidelity. That is a real and useful target. If the proofs hold, logical accreditation is a practical audit path for FT computations and fills a gap in twirling. The extension of entropy benchmarking to the fault-tolerant regime is also a concrete step. Credit where it's due: the problem is well chosen and the proposal connects verification, QEC, and error mitigation in a way that should attract serious readers.\n\nNow the soft spots. The load-bearing assertion is the conversion of arbitrary noise to stochastic Pauli noise, including non-transversal gates. That is exactly where I'd expect trouble—coherent and temporally correlated errors don't always average out. The abstract doesn't show me the construction, and the full text I received is corrupted mojibake (it even cites a different arXiv paper header), so I cannot check whether the bound is conditional on a noise model fitted from the runs being certified. I also see no code or data for the numerical crossover claims. Those are not accusations; they are 'unverified.' Nothing in the abstract makes me think the work is sloppy, but I can't confirm the central theorem.\n\nWho gets value: verification and QEC researchers, plus anyone planning error-mitigation pipelines for logical qubits. If the twirling construction is sound, this is a publishable result worth talking about. As it stands, it's an abstract with high stakes and no inspectable proof. I'd recommend a serious referee rather than a desk reject, but only after the authors supply a readable manuscript with the twirling theorem, the infidelity bound, and simulation details. I'd also ask for code or data release; that would make the numerical claims checkable.","headline":"High-stakes, plausible claims about certifying fault-tolerant computations, but the only readable part is the abstract; send it to a referee after the authors supply a clean manuscript.","tokens_in":21734,"tokens_out":2544,"would_cite":false,"duration_ms":26957,"reading_group":"maybe","serious_thinker":"unclear","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"Logical accreditation claims that randomized compilation can turn arbitrary logical circuit noise into stochastic Pauli noise, making fault-tolerant computations certifiable without classical simulation.","keywords":["logical accreditation","fault-tolerant quantum computation","randomized compiling","Pauli twirling","non-transversal gates","quantum certification","infidelity bounds","entropy benchmarking"],"falsifier":"Run a logical circuit with a deliberately injected coherent error—for example, a small systematic rotation about the $Z$ axis on a logical qubit—through the compiled protocol. If the framework's claim is correct, the certificate must upper-bound the actual infidelity of the output. If the certificate falls below the measured infidelity, or if the Pauli error rates estimated from test runs fail to predict the error on a target run under a known non-Pauli injection, the core conversion-to-Pauli claim is disproved.","tokens_in":20884,"feed_emoji":"⚛️","tokens_out":7831,"duration_ms":76385,"temperature":0.7,"pith_summary":"The paper attempts to close a gap that opens as fault-tolerant quantum computers grow: the outputs of logical circuits become too complex to verify by classical simulation, so certification must be done another way. It introduces logical accreditation, a protocol that randomly recompiles each logical circuit so that the effective noise becomes stochastic Pauli noise—random discrete bit- and phase-flip errors—rather than arbitrary coherent or correlated noise. This twirling works even for non-transversal logical gates beyond the standard T gate, which has been an open problem. Once the noise is Pauli, test runs supply estimates of the error rates, and the protocol uses them to upper bound the infidelity of the logical output state with polynomial overhead. Numerical simulations are reported showing the method can certify quantum advantage experiments and identify the crossover point at which encoded logical computation outperforms physical computation.","feed_headline":"Twirling turns logical noise into stochastic Pauli noise","feed_subtitle":"Logical accreditation certifies fault-tolerant runs without classical simulation, at polynomial overhead.","key_machinery":"The load-bearing mechanism is randomized compilation by Pauli twirling: each logical gate is conjugated by randomly chosen Pauli operations, and averaging over the random choices makes the effective noise a stochastic Pauli channel instead of an arbitrary quantum channel. The new element is a twirling method for non-transversal logical gates beyond the standard T gate, which previous work had left open. This machinery is what turns an intractable certification problem into a statistical one—Pauli error rates can be sampled from test circuits and then used to bound the error in the computation actually being certified.","core_discovery":"The central claim is that arbitrary noise on a logical circuit can be converted into stochastic Pauli noise by a randomized compilation scheme. The scheme includes a way to twirl non-transversal logical gates beyond the standard T gate, resolving an open problem in Pauli twirling of logical gate sets. The Pauli error rates, estimated from random test runs, then yield a rigorous upper bound on the infidelity of the logical output state of the target computation. The authors further claim that the protocol is robust to general noise models far beyond the usual assumptions of quantum error-correction analyses, that it scalably certifies quantum advantage experiments, that it indicates the cross","pith_inferences":["Beyond the paper, the same test-run/target-run design could be packaged as a per-job certification service on cloud quantum processors: every submitted logical computation ships with randomized compiled variants and a returned infidelity certificate.","If coherent noise is genuinely twirled away, then discrepancies between the Pauli-model prediction and measured output errors would become a practical diagnostic for non-Pauli or time-correlated noise, giving a direct experimental test of the framework's core premise.","The twirling technique might transfer to other encoded computing paradigms, such as measurement-based or fusion-based implementations, wherever logical gates admit Pauli conjugation; this extension is not developed in the paper.","The crossover point between logical and physical computation could serve as a single-number hardware benchmark for fault-tolerance roadmaps, since it summarizes when the overhead of encoding starts paying for itself."],"forward_implications":["Fault-tolerant computations can be certified without classically simulating the target run, with overhead that scales polynomially in the computation size.","Certification holds under general noise models, not only the standard assumptions used in quantum error-correction analyses.","The framework can certify quantum advantage experiments and locate the crossover point where encoded logical computation begins to outperform physical computation.","It provides a criterion for when logical error rates are low enough that error mitigation is efficient, and extends entropy benchmarking to the fault-tolerant regime.","A numerical upper bound on the infidelity of the logical output state is produced for each certified computation."],"supporting_citations":[{"why":"Posed the open problem of twirling non-transversal logical gates beyond the standard T gate; the paper's compilation scheme resolves this problem and builds on it.","marker":"[Piveteau et al. PRL 127, 200505 (2021)]"}],"fun_headline_variants":["Turn logical noise into Pauli noise with random twirls","Certify fault-tolerant runs without classical simulation","Logical accreditation: scalable certification for quantum advantage","Twirl non-transversal gates to certify logical output","Resolves open twirling problem for logical gates"],"cache_read_input_tokens":2816,"weakest_assumption_plain":"The framework stands on the premise that the randomized compilation genuinely turns all logical-circuit noise—including noise on non-transversal gates—into stochastic Pauli noise, and that error rates measured on test runs statistically control the error on the target run; if coherent or time-correlated noise survives compilation, the infidelity bound and the scalability claim do not follow.","fun_headline_variants_meta":{"raw":{"variants":["Turn logical noise into Pauli noise with random twirls","Certify fault-tolerant runs without classical simulation","Logical accreditation: scalable certification for quantum advantage","Twirl non-transversal gates to certify logical output","Resolves open twirling problem for logical gates"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000229,"raw_usage":{"total_tokens":1317,"prompt_tokens":747,"completion_tokens":570,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":491,"completion_tokens_details":{"reasoning_tokens":497}},"tokens_in":491,"tokens_out":570,"duration_ms":6153,"temperature":1.0,"reasoning_tokens":497,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-05T23:16:06.942069+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Run a logical circuit with a deliberately injected coherent error—for example, a small systematic rotation about the $Z$ axis on a logical qubit—through the compiled protocol. If the framework's claim is correct, the certificate must upper-bound the actual infidelity of the output. If the certificate falls below the measured infidelity, or if the Pauli error rates estimated from test runs fail to predict the error on a target run under a known non-Pauli injection, the core conversion-to-Pauli claim is disproved.","supporting_citations":[],"review_version":1}