{"id":"0f3eb262-827a-48bb-af88-2b03c2110d05","arxiv_id":"2606.00932","paper_version":1,"verdict":"UNVERDICTED","confidence":"LOW","novelty_score":7.0,"correctness_risk":"unknown","formal_verification":"none","parameter_count":0,"one_line_summary":"Active quantum subspaces enable hybrid quantum advantage by encoding only subsets of data, with proven positive-semidefinite kernels, improvement criteria, and PAC bounds that hold under polynomial encoding cost.","lead":"This paper introduces active quantum subspace data-encoding, where only an information-bearing subset of classical input is lifted to quantum representation while the rest stays classical. It proves structural properties of the resulting hybrid model and shows that polynomial encoding cost need not eliminate hybrid learning advantage.","discovery_kind":"new_method","skeptic_critique":{"model":"grok-4.3","headline":"The nec+suff criterion for hybrid improvement is proved but not verified to hold for the Clifford family under local dephasing noise.","rationale":"The reader's weakest_assumption directly identifies the load-bearing gap between the proved structural criterion and its application to the noisy Clifford construction. The full text supplies the three structural results and the reliability calculation, but the concrete verification step for the correlation condition under noise is missing, so the headline claim remains conditional on that step holding.","tokens_in":1755,"tokens_out":378,"duration_ms":38007,"concrete_test":"From the explicit construction of the 64-qubit Clifford family and synthetic task, extract the projected quantum observable and the classical feature vectors; on a fresh test set of 2000 samples, fit the classical predictor, compute the residual, and evaluate the absolute Pearson correlation between the quantum feature values and that residual. If the correlation is statistically indistinguishable from zero (p > 0.05), the improvement criterion fails.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The central claim that polynomial encoding cost does not destroy the hybrid learning advantage requires that the projected quantum sector actually improves over the classical predictor. The paper states a necessary and sufficient condition for this in squared loss: the projected quantum sector must contain a direction outside the classical feature span that correlates with the classical residual. It then shows oracle reliability remains inverse-polynomial for the Clifford active-subspace family. However, reliability concerns only the noisy oracle; it does not establish that the quantum direction remains orthogonal to the classical span or retains nonzero correlation with the residual once the classical predictor is fitted. The 64-qubit synthetic task illustrates compression of a high-order interaction but supplies no explicit inner-product or residual-correlation calculation under dephasing, leaving open the possibility that noise renders the quantum sector redundant with classical features.","agreement_with_reader":"agree"},"referee_report":{"model":"grok-4.3","summary":"The manuscript introduces active quantum subspace data-encoding, in which only an information-bearing subset of classical inputs is lifted to a quantum representation. It defines a projected hybrid readout and proves three structural results: (1) the projected hybrid kernel is positive semidefinite with sample-regularized dimension bounded by the number of projected observables; (2) a necessary and sufficient criterion for squared-loss improvement over a purely classical predictor (the projected quantum sector must contain a direction outside the classical feature span that correlates with the classical residual); (3) a PAC sample-complexity bound proportional to the inverse square of oracle reliability. For a canonical Clifford active-subspace family under local dephasing, oracle reliability remains inverse-polynomial despite polynomial encoding-gate complexity, from which the authors conclude that polynomial encoding cost does not destroy the hybrid learning advantage. The claim is illustrated by a 64-qubit synthetic contextual classification task.","tokens_in":1907,"tokens_out":550,"duration_ms":16037,"significance":"If the unverified step holds, the work supplies a concrete, QRAM-free route to scalable hybrid quantum advantage on NISQ hardware by avoiding both full data encoding and kernel-dimension blow-up. The explicit nec+suff criterion and the inverse-polynomial reliability result under noise are technically useful even if the specific family requires further checks.","major_comments":[{"comment":"Abstract and the paragraph deriving the Clifford-family reliability bound: the central claim that 'the polynomial encoding cost does not by itself destroy the hybrid learning advantage' requires that the projected quantum sector satisfy the nec+suff criterion (outside classical span + nonzero residual correlation) under local dephasing. The manuscript establishes inverse-polynomial oracle reliability but supplies no explicit inner-product calculation or residual-correlation verification for the dephased Clifford family, leaving open the possibility that noise renders the quantum direction redundant.","section":"Abstract / Clifford-family reliability paragraph"},{"comment":"64-qubit synthetic task illustration: the example demonstrates compression of a high-order interaction into a low-dimensional hybrid model but reports no numerical values for the inner product between the projected quantum feature and the classical residual (or the orthogonality to the classical span) under the stated dephasing noise, so it does not confirm that the nec+suff condition is met in the concrete instance.","section":"64-qubit illustration paragraph"}],"minor_comments":[{"comment":"The abstract is information-dense; expanding the three structural results into a short enumerated list would improve readability for readers who do not yet know the projected-hybrid-kernel construction.","section":"Abstract"}],"recommendation":"major_revision","confidential_remarks":null},"author_rebuttal":{"model":"grok-4.3","summary":"We thank the referee for the careful reading and for highlighting the need for explicit verification of the necessary-and-sufficient improvement criterion under noise. We agree that the central claim requires confirming that the projected quantum sector remains useful (outside the classical span and correlated with the residual) for the dephased Clifford family. Below we respond to each major comment and indicate the revisions we will make.","responses":[{"response":"We agree that the manuscript derives the inverse-polynomial reliability bound for the Clifford family but does not supply the explicit inner-product calculation confirming that the projected quantum direction lies outside the classical span and retains nonzero correlation with the classical residual under local dephasing. This verification is required to close the argument that the reliability bound implies a persistent hybrid advantage. In the revised manuscript we will add a dedicated subsection (or appendix) performing this calculation for the canonical Clifford active-subspace family, showing that the relevant inner product remains bounded away from zero by an inverse-polynomial factor under the stated dephasing model.","revision_made":"yes","referee_comment":"[Abstract / Clifford-family reliability paragraph] Abstract and the paragraph deriving the Clifford-family reliability bound: the central claim that 'the polynomial encoding cost does not by itself destroy the hybrid learning advantage' requires that the projected quantum sector satisfy the nec+suff criterion (outside classical span + nonzero residual correlation) under local dephasing. The manuscript establishes inverse-polynomial oracle reliability but supplies no explicit inner-product calculation or residual-correlation verification for the dephased Clifford family, leaving open the possibility that noise renders the quantum direction redundant."},{"response":"We agree that the 64-qubit synthetic illustration would be strengthened by reporting the numerical values of the inner product between the projected quantum feature and the classical residual (and the distance to the classical span) under the dephasing noise model. The current text only demonstrates compression of the interaction; it does not numerically confirm the nec+suff condition. In the revision we will augment the illustration paragraph (and the associated figure caption or table) with these explicit numerical values computed for the dephased 64-qubit instance, thereby verifying that the condition holds in the concrete example.","revision_made":"yes","referee_comment":"[64-qubit illustration paragraph] 64-qubit synthetic task illustration: the example demonstrates compression of a high-order interaction into a low-dimensional hybrid model but reports no numerical values for the inner product between the projected quantum feature and the classical residual (or the orthogonality to the classical span) under the stated dephasing noise, so it does not confirm that the nec+suff condition is met in the concrete instance."}],"tokens_in":1549,"tokens_out":566,"duration_ms":16542,"standing_objections":[]},"desk_editor":{"model":"grok-4.3","letter":"The main thing here is that they formalize active quantum subspaces where only part of the input goes quantum, prove the hybrid kernel is PSD with dimension bounded by the number of observables, and give a nec and suff condition for beating classical in squared loss: the quantum direction must be outside classical span and correlate with residual.\n\nThey also give a PAC bound based on oracle reliability and show for Clifford family under local dephasing that reliability can be inverse poly with poly gates. That's useful because it directly addresses the encoding cost issue in hybrid QML.\n\nThe 64 qubit synthetic task shows how one quantum feature can capture high order interaction.\n\nThe soft spot is that while they prove the condition and the reliability, they don't show the correlation survives the noise in the example. The reliability is about the oracle, not the post-fit usefulness. So the claim that poly cost doesn't destroy advantage is conditional on that direction staying good, which isn't verified numerically.\n\nStill, the math is clean and the results are structural, not just empirical.\n\nThis is for people in quantum machine learning looking at hybrid models. It has enough formal content to deserve referee time, even if the example is illustrative rather than conclusive.\n\nI'd send it to review.","headline":"Active subspaces give a clean nec+suff condition for hybrid advantage and noise-resilient bounds for Clifford encodings, though the numerical check doesn't fully confirm the quantum direction adds independent value under noise.","tokens_in":2396,"tokens_out":336,"would_cite":false,"duration_ms":13832,"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":"Active quantum subspaces let hybrid learning retain advantage without encoding entire inputs into quantum states.","keywords":["active quantum subspace","hybrid quantum-classical learning","projected hybrid readout","quantum data encoding","NISQ advantage","sample complexity","Clifford circuits","local dephasing noise"],"falsifier":"A demonstration that oracle reliability falls faster than inverse-polynomially as the number of encoding gates increases polynomially, or that no performance gain occurs in the synthetic contextual classification task when adding the projected quantum feature.","tokens_in":2658,"feed_emoji":"⚛️","tokens_out":446,"duration_ms":23310,"temperature":0.7,"pith_summary":"The paper examines if quantum learning advantage requires fully embedding large classical data into superposed quantum states. It proposes active quantum subspace encoding, lifting only a useful subset to quantum while keeping others classical. Structural results show the hybrid kernel is positive semidefinite with bounded dimension, and give a criterion for when the quantum part beats a classical predictor. In noisy settings, sample complexity depends on oracle reliability, which stays inverse-polynomial for certain Clifford families even as encoding gates grow polynomially. This indicates that encoding costs alone do not eliminate the hybrid edge.","feed_headline":"Partial quantum encoding keeps hybrid learning advantage alive","feed_subtitle":"Projected subspaces maintain inverse-polynomial reliability under noise even with polynomial gate costs, avoiding full data encoding.","key_machinery":"Active quantum subspace data-encoding with projected hybrid readout, which lifts only an information-bearing subset of inputs to quantum representation.","core_discovery":"In the projected hybrid readout model with active quantum subspace encoding, the projected hybrid kernel is positive semidefinite with sample-regularized dimension bounded by the number of projected observables. A necessary and sufficient condition for squared-loss improvement over classical predictors is that the projected quantum sector contains a direction outside the classical feature span that correlates with the classical residual. In a realizable noisy-oracle setting the PAC sample complexity scales as the inverse square of oracle reliability, and for canonical Clifford active-subspace families under local dephasing this reliability remains inverse-polynomial even when encoding gate c","pith_inferences":[],"forward_implications":[],"fun_headline_variants":["Active subspaces deliver hybrid advantage without full data encoding","Projected kernels bound dimension for scalable hybrid learning","Quantum sector must correlate outside classical span for gains","Noisy Clifford families retain inverse-polynomial reliability at scale"],"cache_read_input_tokens":2112,"weakest_assumption_plain":"The projected quantum sector must contain a direction that lies outside the classical feature span and correlates with the classical residual.","fun_headline_variants_meta":{"raw":{"variants":["Active subspaces deliver hybrid advantage without full data encoding","Projected kernels bound dimension for scalable hybrid learning","Quantum sector must correlate outside classical span for gains","Noisy Clifford families retain inverse-polynomial reliability at scale"]},"model":"grok-4.3","cost_usd":0.005234,"raw_usage":{"total_tokens":2499,"prompt_tokens":757,"num_sources_used":0,"completion_tokens":58,"cost_in_usd_ticks":52340500,"prompt_tokens_details":{"text_tokens":757,"audio_tokens":0,"image_tokens":0,"cached_tokens":64},"completion_tokens_details":{"audio_tokens":0,"reasoning_tokens":1684,"accepted_prediction_tokens":0,"rejected_prediction_tokens":0}},"tokens_in":757,"tokens_out":58,"duration_ms":13109,"temperature":1.0,"reasoning_tokens":1684,"cache_read_input_tokens":64,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-06-28T18:14:17.517925+00:00","model_set":{"reader":"grok-4.3"},"falsifier":"A demonstration that oracle reliability falls faster than inverse-polynomially as the number of encoding gates increases polynomially, or that no performance gain occurs in the synthetic contextual classification task when adding the projected quantum feature.","supporting_citations":[],"review_version":1}