{"id":"428af8e4-feb8-40c7-8d1e-cade5e7703d3","arxiv_id":"2608.04822","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":4.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Under post-selected no-jump filtering, qubit states with a fixed quantum Hamming weight are perfectly preserved against amplitude-damping noise, while superposition states face a fidelity lower bound set by their Hamming distance.","lead":"This paper analyzes a quantum filter circuit that post-selects on the no-jump outcome of amplitude-damping noise, showing that states of fixed Hamming weight can be protected with unit fidelity. It also derives fidelity bounds for superpositions such as GHZ states and compares the approach to controlled-SWAP error mitigation.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Single-ancilla perfect protection does not hold for all fixed Hamming weights: Eq. (18) itself gives F<1 for s>=2, contradicting the abstract and conclusion.","rationale":"The reader's CONDITIONAL verdict identified the input-subspace assumption as the weakest point. My check shows that this condition is necessary but not sufficient: even within a single H_{s,N}, the single-ancilla filter gives F=1 only for s=0 or s=1. The paper itself supplies the formula that disproves the broad claim, so this is an internal inconsistency rather than a disagreement with prior work. Correcting the claim reduces the advertised result to 'weight-1 states can be perfectly protected with one ancilla, and any fixed-weight-s state with N ancillas', which invalidates the abstract and conclusion as written. The reader did notice that Eq. (18) gives F=1 only for s=1 and for any s with N ancillas, but the verdict did not treat the abstract's overgeneralization as fatal. Because the central advertised claim is false as stated, I move the verdict to REJECT; a revised version that restricts the claim to s=1 for single-ancilla protection, or to the N-ancilla setting for general s, would be suitable for re-review.","tokens_in":11480,"tokens_out":6987,"duration_ms":75775,"concrete_test":"Evaluate Eqs. (18)-(19) for s=2, N=2 at gamma=0.2: p0=(0.8)^2+(0.2)^2=0.68 and F=sqrt(0.64/0.68)=0.970<1. Alternatively, run the single-control filter on |11>: E00|11>=(1-gamma)|11> and E11|11>=gamma|00> both survive post-selection, giving rho_out proportional to (1-gamma)^2|11><11|+gamma^2|00><00|, so the fidelity is below 1 for any 0<gamma<1. This directly settles whether all fixed Hamming weights are perfectly protected.","verdict_should_be":"REJECT","load_bearing_attack":"The paper's headline claim that 'states possessing a fixed quantum Hamming weight can be perfectly protected against AD noise using only a single ancilla' (Conclusion; cf. Abstract) is contradicted by its own fidelity formula. For any input in H_{s,N}, the post-selected fidelity is F_{s,N}=sqrt((1-gamma)^s/p0) with p0=sum over even m of C(s,m)(1-gamma)^{s-m}gamma^m (Eqs. 18-19). For s>=2, p0 contains the m=2 term C(s,2)(1-gamma)^{s-2}gamma^2>0, so p0>(1-gamma)^s and F_{s,N}<1. Example: the fixed-weight-2 state |11> passes through E00 to (1-gamma)|11> and through E11 to gamma|00>; the post-selected mixture has fidelity (1-gamma)/sqrt((1-gamma)^2+gamma^2), not 1. Eq. 18 already restricts perfect single-ancilla protection to s=0,1; Section III.A states this correctly for s=1, and Section IV achieves perfect protection for general s only with N control qubits (Eq. 29). The abstract and conclusion therefore overgeneralize the central result. The same issue appears in the Bell-state discussion: the state they call |Phi+-> with unit fidelity is the standard Hamming-weight-1 Bell state |Psi+->, while the standard |Phi+-> has weights 0 and 2 and is not perfectly protected.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper proposes a quantum filter circuit based on controlled-Z gates and post-selection to protect quantum states against amplitude-damping (AD) noise. For a single qubit it shows that the filter removes the E1 damping term and protects |1> with unit fidelity, outperforming a controlled-SWAP circuit. The treatment is extended to Bell states, to multi-qubit states within fixed Hamming-weight subspaces, and to superpositions such as GHZ states. The main analytical results are the post-selected fidelity formula F_{s,N} = sqrt((1-gamma)^s / p0) for the single-ancilla filter (Eq. 18), the claim that states of fixed Hamming weight can be perfectly protected with a single ancilla, and the N-ancilla construction of Section IV that perfectly protects any fixed-weight subspace H_{s,N}. Numerical results for average fidelities and a comparison with CSWAP are also presented.","tokens_in":11769,"tokens_out":10679,"duration_ms":106857,"significance":"If the advertised claims were correct in their stated generality, the paper would offer a striking resource-efficient alternative to full quantum error correction for amplitude-damping noise: a simple CZ filter with post-selection supposedly protects all fixed-weight states with a single ancilla. The paper does provide explicit, internally consistent fidelity formulas, including Eq. (18), and the N-ancilla construction in Section IV is correct. However, the central single-ancilla claim is overstated in the abstract and conclusion, the Bell-state labeling is nonstandard and reverses which states are perfectly protected, and the proof of the GHZ worst-case claim in Appendix B is not valid. These issues affect the paper's main advertised message, so the work needs substantial revision before the claims can be accepted.","major_comments":[{"comment":"","section":"Abstract, Conclusion, Section III.A, Eq. (18)-(19)"},{"comment":"","section":"Section II.B"},{"comment":"","section":"Section IV and Appendix B"}],"minor_comments":[{"comment":"","section":"Appendix A"},{"comment":"","section":"Section III.B, after Eq. (24)"},{"comment":"","section":"Section V heading"},{"comment":"","section":"Appendix A"},{"comment":"","section":"Keywords line"}],"recommendation":"major_revision","confidential_remarks":"The manuscript is a standard theory paper with analytical formulas and numerical plots; no code or data availability statement is provided, which is not unusual for this type of work. The main concern is that the headline claim is overstated in a way that is easily missed because Eq. (18) is correct but its consequences for s>=2 are not stated. The Bell-state labeling issue and the invalid Appendix B proof also need attention. None of these problems appear unfixable, so I recommend major revision rather than rejection."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Short version: the useful core is correct, but the headline claim is not. Eq. (18) gives the single-ancilla post-selected fidelity for a fixed-weight-s state as sqrt((1-gamma)^s / p0), with p0 containing a nonzero m=2 term for s>=2, so F<1 for s>=2. Perfect protection with one ancilla happens only for s=0 and s=1. The abstract and conclusion claim “fixed quantum Hamming weight” can be perfectly protected with a single ancilla, which is an overgeneralization that contradicts the paper’s own formula.\n\nWhat is genuinely good: applying the commutation-derived quantum filter [36,37] to AD noise is a clean application, and the paper systematically catalogs fidelities for single qubits, Bell states, W/Dicke states, GHZ states, and general superpositions. The N-ancilla circuit, being a tensor product of single-qubit filters, does perfectly protect any fixed-weight state, and the fidelity-invariance lemma (Eq. 17) is correct and useful: within a fixed-weight subspace all input states behave identically.\n\nSoft spots, in proportion: (1) The Bell-state section uses a nonstandard labeling, calling (|00>±|11>)/√2 the |Ψ±> states and reserving |Φ±> for weight-1 states. Under the usual convention it is |Ψ±> that have Hamming weight 1 and are perfectly protected, so the “perfect protection of |Φ±>” claim is confusing and, in any case, asserted without derivation (it does follow from Eq. (18) with s=1). (2) Appendix B’s proof that GHZ states are the worst case is sketchy; it relies on a small-gamma expansion and an inequality chain that is not fully justified. It should be tightened or removed. (3) The CSWAP comparison is numerical and lacks a rigorous resource analysis, so the “resource-efficient” claim is only partially supported. (4) The filter circuit itself is prior work; the novelty is the application and classification, which the paper should state more modestly.\n\nThis is a modest but mostly correct contribution with a fixable overclaim. It deserves serious peer review, but the revision should correct the abstract and conclusion, fix the Bell-state notation, and either repair or cut Appendix B.","headline":"Right math, overstated headline: single-ancilla perfect protection holds only for Hamming weights 0 and 1, not all fixed weights as the abstract claims.","tokens_in":12292,"tokens_out":5604,"would_cite":false,"duration_ms":56712,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"A control-qubit filter can perfectly preserve fixed-excitation-count states under amplitude-damping noise, without full quantum error correction.","keywords":["amplitude damping","quantum filter","state protection","quantum Hamming weight","post-selection","decoherence suppression","multi-qubit states","fidelity"],"falsifier":"Run a single round of the filtration circuit on the one-excitation Bell state $(|01\\rangle+|10\\rangle)/\\sqrt{2}$ under tunable amplitude damping at several rates $\\gamma$, and post-select on the control qubit measuring $|0\\rangle$. The paper predicts the output equals the input, so the conditional fidelity is exactly 1 for every $\\gamma$, with success probability $1-\\gamma$; any measured deviation from 1 would falsify the central claim.","tokens_in":11269,"feed_emoji":"🛡️","tokens_out":12452,"duration_ms":133294,"temperature":0.7,"pith_summary":"All real qubits lose energy to their environment, and the dominant low-temperature error is amplitude damping: the excited state $|1\\rangle$ relaxes to $|0\\rangle$ with some rate $\\gamma$. This paper introduces a filtering circuit, a control qubit coupled through controlled-Z gates followed by post-selection, that suppresses exactly that relaxation. It establishes that any $N$-qubit state lying entirely in a subspace of fixed quantum Hamming weight, meaning every computational basis component contains the same number $s$ of $|1\\rangle$s, can be perfectly restored after amplitude damping: the post-selected fidelity is $F_{s,N}=\\sqrt{(1-\\gamma)^s/p_0}$, which equals 1 for weight $s=1$ with one ancilla and for any fixed weight $s$ when one ancilla per qubit is used. For states that superpose different Hamming weights, perfect protection is not achieved with this filter, and the GHZ state, whose components have maximal Hamming distance, sets the lower bound. The practical upshot is a resource-efficient route to protecting memories and entangled states from energy relaxation without invoking full quantum error correction.","feed_headline":"Filter perfectly protects fixed-excitation states from decay","feed_subtitle":"A control-qubit post-selection removes the damping jump, preserving the state without full error correction.","key_machinery":"The load-bearing object is the quantum filtration circuit: one or more control qubits coupled to the data qubits by controlled-Z gates, followed by measurement and post-selection on the $|0\\rangle$ outcome. It exploits the commutation relation $[E_0,Z]=0$ and the anticommutation relation $\\{E_1,Z\\}=0$ for the amplitude-damping Kraus operators $E_0=|0\\rangle\\langle0|+\\sqrt{1-\\gamma}|1\\rangle\\langle1|$ and $E_1=\\sqrt{\\gamma}|0\\rangle\\langle1|$, so the filter removes every Kraus term containing an odd number of $E_1$ factors. The second central object is the quantum Hamming weight, the number of $|1\\rangle$s in a computational-basis string: the subspace $H_{s,N}$ of fixed weight $s$ acts as a code space because any damping term with $m$ jump operators maps it into an orthogonal subspace $H_{s-m,N}$. The fidelity formula $F_{s,N}=\\sqrt{(1-\\gamma)^s/p_0}$ carries the argument, showing that protection depends only on $s$ rather than on the total qubit number $N$, and becomes unity when only the $m=0$ term contributes.","core_discovery":"The central claim is that for amplitude-damping noise with rate $\\gamma$, the quantum filter preserves the no-jump branch exactly and discards the jump branch. For any input state supported on the $N$-qubit subspace $H_{s,N}$ of fixed quantum Hamming weight $s$, the post-selected output is proportional to the original state and the fidelity is $F_{s,N}=\\sqrt{(1-\\gamma)^s/p_0}$, with $p_0=\\sum_{m\\,\\mathrm{even}} \\binom{s}{m}(1-\\gamma)^{s-m}\\gamma^m$. Because every Kraus operator containing an odd number of $E_1$ factors is filtered out, and every even-$m$ Kraus operator with $m>0$ maps $H_{s,N}$ into an orthogonal subspace, only the no-jump term survives; this yields unit fidelity exactly for weight $s=1$ with a single ancilla and for arbitrary fixed $s$ when the filter uses $N$ ancillas, one per qubit. The paper thus proves that one-excitation Bell states, generalized W states, and Dicke states, all of which have fixed Hamming weight, are perfectly protected, while GHZ states, with maximal Hamming distance between their components, are the worst case and bound the achievable fidelity from below.","pith_inferences":["Editorial inference: because the perfect protection is conditional on a successful post-selection whose probability is $(1-\\gamma)^s$ for an $s$-excitation state, the scheme is a heralded error-detection protocol: it protects accepted runs only, and the unconditional output fidelity is no better than the bare no-jump survival. The paper's formulas imply this trade-off but do not state it.","Editorial inference: the Hamming-weight result suggests a design principle for $T_1$-limited hardware: encode logical information inside a single excitation-number subspace so that damping acts as a detectable leakage event. A natural extension would be to combine the filter with a recovery step on the rejected runs, converting the heralded protection into an erasure-correcting code.","Editorial inference: the same controlled-Z filtration should suppress any noise channel whose Kraus operators split into commuting and anticommuting sets under a Pauli symmetry; amplitude damping is one instance, and the fixed-Hamming-weight protection would then be a special case of symmetry-based error detection."],"forward_implications":["Single-ancilla filtration recovers the excited single-qubit state $|1\\rangle$ and every one-excitation Bell state with unit fidelity after post-selection, at success probability $1-\\gamma$.","With one ancilla per data qubit, every state of fixed Hamming weight $s$ is recovered with unit fidelity, at success probability $(1-\\gamma)^s$, including generalized W states and Dicke states.","Superposition states cannot be perfectly protected: the post-selected fidelity lies strictly between the fidelities of their component Hamming-weight sectors, and the GHZ state, with maximal Hamming distance $N$, is the worst case.","The fidelity depends only on the Hamming weight $s$ and not on the total qubit number $N$, so protection quality within a fixed weight sector does not degrade as the system grows.","The filter outperforms controlled-SWAP-based purification in fidelity and avoids the exponential copy overhead of streaming purification schemes."],"supporting_citations":[{"why":"Supplies the quantum-filter circuit structure that the paper adapts for amplitude-damping noise.","marker":"[36]"},{"why":"Provides the commutation-derived quantum filter idea that motivates the controlled-Z construction.","marker":"[37]"},{"why":"Standard reference for the amplitude-damping channel and its Kraus operators, used to set up the noise model and fidelity.","marker":"[14]"},{"why":"Establishes amplitude damping as the dominant relaxation model and supplies known bosonic codes for it, motivating the protection target.","marker":"[15]"},{"why":"Introduces the quantum Hamming weight and projection circuits used to classify input states by excitation number.","marker":"[45]"},{"why":"Supplies generalized W states as the fixed-Hamming-weight example that achieves perfect protection.","marker":"[34]"},{"why":"Supplies Dicke states as the symmetric fixed-weight example covered by the fidelity formula.","marker":"[35]"},{"why":"Baseline for streaming purification resource overhead that the filter avoids.","marker":"[43]"}],"fun_headline_variants":["Quantum filter perfectly preserves fixed-weight states","Post-selection filter gives unit fidelity for fixed Hamming weight states","Filter alone protects fixed-weight states without full error correction","Fixed-weight states survive decay exactly via filter"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The perfect-protection result holds only for input states that lie entirely within a single quantum Hamming-weight subspace; once a state superposes different Hamming weights, the surviving even-m damping terms create components in other weight sectors and the fidelity drops.","fun_headline_variants_meta":{"raw":{"variants":["Quantum filter perfectly preserves fixed-weight states","Post-selection filter gives unit fidelity for fixed Hamming weight states","Filter alone protects fixed-weight states without full error correction","Fixed-weight states survive decay exactly via filter"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000857,"raw_usage":{"total_tokens":3717,"prompt_tokens":936,"completion_tokens":2781,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":552,"completion_tokens_details":{"reasoning_tokens":2720}},"tokens_in":552,"tokens_out":2781,"duration_ms":22961,"temperature":1.0,"reasoning_tokens":2720,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-06T15:43:12.095151+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Run a single round of the filtration circuit on the one-excitation Bell state $(|01\\rangle+|10\\rangle)/\\sqrt{2}$ under tunable amplitude damping at several rates $\\gamma$, and post-select on the control qubit measuring $|0\\rangle$. The paper predicts the output equals the input, so the conditional fidelity is exactly 1 for every $\\gamma$, with success probability $1-\\gamma$; any measured deviation from 1 would falsify the central claim.","supporting_citations":[{"cited_title":"Greenberger-horne-zeilinger versuswstates: Quantum teleportation through noisy channels,","cited_arxiv_id":null,"evidence_quote":"Supplies generalized W states as the fixed-Hamming-weight example that achieves perfect protection."},{"cited_title":"Error suppression for arbitrary-size black box quantum operations,","cited_arxiv_id":null,"evidence_quote":"Supplies the quantum-filter circuit structure that the paper adapts for amplitude-damping noise."},{"cited_title":"Nielsen and Isaac L","cited_arxiv_id":null,"evidence_quote":"Standard reference for the amplitude-damping channel and its Kraus operators, used to set up the noise model and fidelity."},{"cited_title":"Bosonic quantum codes for amplitude damping,","cited_arxiv_id":null,"evidence_quote":"Establishes amplitude damping as the dominant relaxation model and supplies known bosonic codes for it, motivating the protection target."},{"cited_title":"Logarithmic-depth quantum circuits for hamming weight projections,","cited_arxiv_id":null,"evidence_quote":"Introduces the quantum Hamming weight and projection circuits used to classify input states by excitation number."},{"cited_title":"Symmetric quantum states: a review of recent progress,","cited_arxiv_id":null,"evidence_quote":"Supplies Dicke states as the symmetric fixed-weight example covered by the fidelity formula."},{"cited_title":"Streaming quantum state purification,","cited_arxiv_id":null,"evidence_quote":"Baseline for streaming purification resource overhead that the filter avoids."}],"review_version":1}