{"id":"9f28c9a7-1abf-4e4c-b765-828fafc6fad9","arxiv_id":"2606.04186","paper_version":1,"verdict":"UNVERDICTED","confidence":"LOW","novelty_score":6.0,"correctness_risk":"unknown","formal_verification":"none","parameter_count":2,"one_line_summary":"QFlow algorithm enables hybrid quantum-classical simulation of large orbital spaces in quantum chemistry, recovering over 95% of CCSD correlation energy with the equivalent of 12 qubits for systems up to 114 orbitals.","lead":"The paper introduces the Quantum Flow (QFlow) algorithm, a hybrid quantum-classical method for simulating correlated many-body systems in quantum chemistry using a singles-and-doubles model. A smart generalist might read it to see how limited quantum resources combined with classical computing could enable larger molecular simulations than pure quantum approaches currently allow.","discovery_kind":"new_method","skeptic_critique":{"model":"grok-4.3","headline":"95% CCSD recovery claim hinges on whether 6e/6o active space plus QFlow flow actually captures dynamical correlation without hidden adjustments","rationale":"The reader's weakest_assumption directly isolates the same active-space sufficiency issue that underpins the 95% recovery number. Because the full manuscript was not supplied to the first reader, the present pass cannot locate a more technical flaw (e.g., an inconsistent equation) but confirms that the accuracy claim remains the load-bearing point.","tokens_in":1743,"tokens_out":368,"duration_ms":17330,"concrete_test":"Extract the precise definition of recovered correlation energy (E_QFlow - E_HF) / (E_CCSD - E_HF) from the methods or results section; recompute the percentage after subtracting any post-hoc classical corrections or after restricting CCSD to the same (6,6) active space; if the ratio drops below 80% the central accuracy claim does not hold.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The headline result states that a singles-and-doubles QFlow model with an active space of only 6 electrons in 6 orbitals recovers >95% of the CCSD correlation energy for 82- and 114-orbital systems dominated by dynamical correlation. Dynamical correlation is recovered in CCSD by excitations involving the full virtual space; restricting the quantum active space to (6,6) while claiming near-complete recovery therefore requires that the parallel classical flow exactly compensates for all missing dynamical contributions. No explicit partitioning of the correlation energy into active-space versus external contributions, nor any demonstration that the flow equations preserve size-consistency or avoid double-counting, is visible in the provided abstract. This makes the numerical percentage the least-secured step in the argument.","agreement_with_reader":"agree"},"referee_report":{"model":"grok-4.3","summary":"The manuscript introduces the Quantum Flow (QFlow) algorithm, a hybrid quantum-classical framework for correlated many-body systems. It reports an HPC implementation of a singles-and-doubles model applied to 82- and 114-orbital systems using a (6 electrons, 6 orbitals) active space. The work claims that this setup optimizes 1.17 million wave-function parameters with the equivalent of 12 qubits while recovering over 95% of the CCSD correlation energy for systems dominated by dynamical correlation, and that the approach remains accurate in extended basis sets with diffuse functions.","tokens_in":1918,"tokens_out":484,"duration_ms":15804,"significance":"If substantiated, the result would demonstrate a practical route to combining limited quantum resources for active-space correlation with classical flow for external dynamical contributions, potentially extending the reach of quantum algorithms to larger molecular systems where full CCSD remains expensive. The reported scale of parameter optimization (1.17 million parameters) with modest qubit counts is a concrete technical achievement worth noting.","major_comments":[{"comment":"Abstract: The headline claim of recovering >95% of the total CCSD correlation energy with a (6e,6o) active space for 82- and 114-orbital systems is load-bearing. No explicit partitioning of the correlation energy into active-space versus external contributions, nor any demonstration that the flow equations preserve size-consistency or avoid double-counting of dynamical correlation, is visible; without this, it is unclear how the classical component exactly compensates for the excitations omitted from the small active space.","section":"Abstract"},{"comment":"Abstract: The assertion that dynamical correlation effects 'remain challenging for existing quantum algorithms' is used to position the result, yet the manuscript provides no direct numerical comparison to other hybrid active-space or embedding methods on the same 82- and 114-orbital test cases, making the relative performance gain difficult to quantify.","section":"Abstract"}],"minor_comments":[{"comment":"The abstract refers to an 'HPC implementation' and 'high-performance computing' but supplies no concrete details on wall-clock time, node count, or parallel scaling; these would strengthen the resource-efficiency claim.","section":"Abstract"}],"recommendation":"major_revision","confidential_remarks":null},"author_rebuttal":{"model":"grok-4.3","summary":"We thank the referee for the detailed and constructive report. We address the two major comments point-by-point below. Where the comments identify opportunities for clarification, we indicate the revisions that will be made in the resubmitted manuscript.","responses":[{"response":"The QFlow construction separates the problem by design: the quantum solver computes the exact active-space (6e,6o) amplitudes, while the classical flow equations propagate these amplitudes into the external orbital space via a singles-and-doubles truncation that adds only the missing dynamical contributions. Because the flow operates on the residual amplitudes outside the active space, the active-space correlation is not re-counted. Size consistency follows from the additive structure of the flow equations under the same truncation used in CCSD. We agree that an explicit statement of this partitioning and a short derivation confirming size consistency and absence of double-counting would strengthen the abstract and introduction; these clarifications will be added in the revised manuscript.","revision_made":"partial","referee_comment":"[Abstract] Abstract: The headline claim of recovering >95% of the total CCSD correlation energy with a (6e,6o) active space for 82- and 114-orbital systems is load-bearing. No explicit partitioning of the correlation energy into active-space versus external contributions, nor any demonstration that the flow equations preserve size-consistency or avoid double-counting of dynamical correlation, is visible; without this, it is unclear how the classical component exactly compensates for the excitations omitted from the small active space."},{"response":"The positioning statement reflects the well-documented resource scaling of full CCSD on systems of this size, which exceeds near-term quantum hardware. While we do not provide head-to-head benchmarks against other hybrid or embedding schemes on these exact molecules (such comparisons would require substantial additional implementation outside the scope of the present Letter), the reported scale—optimization of 1.17 million parameters with the equivalent of 12 qubits—illustrates a concrete resource advantage. A concise discussion of related hybrid approaches will be inserted in the introduction to better contextualize the result.","revision_made":"partial","referee_comment":"[Abstract] Abstract: The assertion that dynamical correlation effects 'remain challenging for existing quantum algorithms' is used to position the result, yet the manuscript provides no direct numerical comparison to other hybrid active-space or embedding methods on the same 82- and 114-orbital test cases, making the relative performance gain difficult to quantify."}],"tokens_in":1393,"tokens_out":495,"duration_ms":15268,"standing_objections":[]},"desk_editor":{"model":"grok-4.3","letter":"Colleague,\n\nThe main takeaway is that this paper introduces a hybrid QFlow method that runs a small active-space problem on limited quantum resources while a classical flow supposedly recovers most of the dynamical correlation. They report optimizing 1.17 million parameters with the equivalent of 12 qubits and getting over 95% of CCSD correlation energy on 82- and 114-orbital systems.\n\nWhat is new is the QFlow formalism itself and its HPC implementation for a singles-and-doubles model. The tests on extended basis sets with diffuse functions and the scale of the orbital spaces are concrete steps beyond typical small-molecule quantum algorithm demos. If the parallel resource split works without extra fitting, it could interest people looking for ways to stretch current hardware.\n\nThe soft spot is the 95% recovery claim. Dynamical correlation in CCSD requires the full virtual space, yet the quantum part is restricted to 6 electrons in 6 orbitals. The abstract does not show how the flow equations separate active-space and external contributions, whether size consistency holds, or any error breakdown. That leaves the headline percentage as the least-supported part of the argument, exactly as the stress-test note flags.\n\nThis is for researchers working on hybrid quantum-classical algorithms for chemistry who want to see resource-efficient scaling ideas. A reader already following variational or embedding methods might get something out of the framework even if the numbers need checking.\n\nI would send it to peer review. The implementation claims are specific enough that referees can test whether the flow actually delivers the reported accuracy without hidden adjustments.","headline":"QFlow claims 95% CCSD recovery on 82-114 orbital systems from a (6,6) active space plus classical flow, but the abstract gives no partitioning or consistency checks to back the number.","tokens_in":2411,"tokens_out":405,"would_cite":false,"duration_ms":15970,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"grok-4.3","headline":"The Quantum Flow algorithm recovers over 95 percent of CCSD correlation energy using the equivalent of only 12 qubits for systems with up to 114 orbitals.","keywords":["quantum flow algorithm","hybrid quantum-classical","correlation energy","quantum chemistry","active space","dynamical correlation","CCSD recovery"],"falsifier":"A direct numerical comparison on a system with stronger static correlation or a larger active space where the recovered fraction of CCSD energy drops well below 95 percent.","tokens_in":2645,"feed_emoji":"⚛","tokens_out":630,"duration_ms":16241,"temperature":0.7,"pith_summary":"The paper presents a high-performance computing implementation of the Quantum Flow (QFlow) formalism in a singles-and-doubles model for hybrid quantum-classical architectures. It shows that this approach can handle target spaces of 82 and 114 orbitals by including all six active electrons in six active orbitals while optimizing over a million wave function parameters. The key result is that QFlow recovers more than 95 percent of the correlation energy obtained from coupled cluster singles and doubles calculations for systems dominated by dynamical correlation. These systems remain difficult for existing quantum algorithms despite the modest qubit requirements. The formalism also preserves accuracy when using extended basis sets that include diffuse functions.","feed_headline":"QFlow recovers 95% of CCSD correlation energy with 12 qubits","feed_subtitle":"Hybrid method optimizes 1.17 million parameters for 114-orbital systems while using modest quantum resources for dynamical correlation.","key_machinery":"The Quantum Flow (QFlow) formalism that enables parallel utilization of quantum and classical resources to describe correlated many-body systems on hybrid architectures.","core_discovery":"The singles-and-doubles QFlow model, run on hybrid quantum-classical hardware, optimizes 1.17 million wave function parameters with the equivalent of 12 qubits and recovers over 95 percent of the total correlation energy from CCSD for 82- and 114-orbital systems dominated by dynamical correlation effects.","pith_inferences":["Similar parallel harvesting strategies might be adapted to other many-body simulation domains that mix quantum and classical resources.","The method could serve as an intermediate step between current small-qubit devices and full-scale quantum simulations of chemistry.","Testing on molecules with known experimental energies would clarify how close the 95 percent figure comes to chemical accuracy."],"forward_implications":["The approach offers a scalable route to quantum chemistry simulations of realistic molecular systems.","Accuracy holds in extended basis sets containing diffuse functions.","It targets dynamical correlation regimes that current quantum algorithms handle poorly.","The hybrid parallel structure reduces qubit demands while still optimizing over a million parameters."],"fun_headline_variants":["QFlow recovers 95% CCSD correlation with 12 qubits in 114 orbitals","QFlow optimizes 1.17M parameters with 12 qubits for 114 orbitals","95% CCSD correlation recovered using QFlow and 12 qubits","QFlow uses 12 qubits for 95% CCSD recovery in 114 orbitals"],"cache_read_input_tokens":2112,"weakest_assumption_plain":"The singles-and-doubles QFlow model restricted to a small 6-electron-in-6-orbital active space captures enough of the correlation energy to reach the claimed 95 percent recovery without further corrections.","fun_headline_variants_meta":{"raw":{"variants":["QFlow recovers 95% CCSD correlation with 12 qubits in 114 orbitals","QFlow optimizes 1.17M parameters with 12 qubits for 114 orbitals","95% CCSD correlation recovered using QFlow and 12 qubits","QFlow uses 12 qubits for 95% CCSD recovery in 114 orbitals"]},"model":"grok-4.3","cost_usd":0.013295,"raw_usage":{"total_tokens":5741,"prompt_tokens":631,"num_sources_used":0,"completion_tokens":85,"cost_in_usd_ticks":132949500,"prompt_tokens_details":{"text_tokens":631,"audio_tokens":0,"image_tokens":0,"cached_tokens":256},"completion_tokens_details":{"audio_tokens":0,"reasoning_tokens":5025,"accepted_prediction_tokens":0,"rejected_prediction_tokens":0}},"tokens_in":631,"tokens_out":85,"duration_ms":33920,"temperature":1.0,"reasoning_tokens":5025,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-06-28T09:28:32.833789+00:00","model_set":{"reader":"grok-4.3"},"falsifier":"A direct numerical comparison on a system with stronger static correlation or a larger active space where the recovered fraction of CCSD energy drops well below 95 percent.","supporting_citations":[],"review_version":1}