{"id":"af388f1c-660e-4b76-9763-169d5391d45f","arxiv_id":"2608.02297","paper_version":1,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":7.0,"correctness_risk":"low","formal_verification":"none","parameter_count":0,"one_line_summary":"Under quantum resetting with uniform memory, every density-matrix element evolves exactly as a Kummer function of its Bohr frequency, producing algebraic decoherence in gapped systems and logarithmically slow spreading in gapless systems.","lead":"This paper introduces a quantum version of 'reset to a randomly remembered past state', where each reset sends the system to a state it occupied at a uniformly random earlier time. It derives exact equations for how such memory-reset quantum systems evolve, showing that they slow down dramatically and either freeze into a memory-dependent steady state or spread extremely slowly, depending on the energy spectrum.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Universal gapless scaling (Eq. 49) rests on an uncontrolled joint limit in Appendix C; without an error bound on off-diagonal initial coherences, the claimed log(rt)/r spreading independent of initial state is not fully established.","rationale":"I read the paper in good faith and checked the central derivation. The exact solution Eq. (13) is clean: multiplying Eq. (9) by t, differentiating, and mapping to Kummer's equation is valid; the initial conditions f(0)=1 and f'(0)=-iω are correct; and the DLMF asymptotic (14) is the appropriate leading term because Re[-(r+iω)t] → -∞. The gapped-sector conclusion (diagonal elements fixed, off-diagonals decaying algebraically, stationary state independent of r) follows rigorously for Δ>0. The reader's weakest assumption is exactly the place where I also see the main gap: Appendix C derives the universal log(rt)/r scaling by a joint limit with no error estimates. The approximation f(ω,t) ≈ e^{-iω log(rt)/r} is formally correct for ω ~ 1/log t, and the diagonal replacement is plausible for smooth initial density matrices, but the paper gives no dominated-convergence argument or bound on the remainder. In particular, the abstract's phrase 'independently of the initial state' overstates the conditionality in Appendix C. This is a genuine but nonfatal concern: a concrete numerical test using the exact f would settle whether any physical initial state actually violates the predicted collapse. Since the reader already assigned CONDITIONAL, my concern does not change the verdict.","tokens_in":23130,"tokens_out":17907,"duration_ms":151328,"concrete_test":"For a fixed initial state with strong off-diagonal momentum coherences (e.g., a Schrödinger-cat superposition of two Gaussian wave packets separated by δ in momentum), evaluate the exact position distribution from Eq. (C.5) using the full hypergeometric f(ω,t) in Eq. (12), with no further approximation, at t = 10^6, 10^8, 10^10 for several r. Plot r P_r(x,t)/log(rt) versus v = rx/log(rt) and compare to Eq. (49). If the curves collapse to V0(v) with deviations decaying faster than 1/log(rt), the joint limit lands; if the collapse fails for some δ or the width does not grow as log(rt)/r, the universality claim must be restricted.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The exact solution Eq. (13) and the gapped-sector results are solid: the ODE derivation, the Kummer function identification, and the long-time asymptotics are consistent, and the two-level Monte Carlo supports Eq. (32). The load-bearing soft spot is the gapless scaling derivation in Appendix C. The transition from the exact Eq. (C.5) to Eq. (C.10) uses two approximations: (i) f(qv(p),t) ≈ exp(-i q v(p) log(rt)/r) in the joint limit q ~ 1/log(rt), and (ii) ρ0(p+q/2, p-q/2, 0) → ρ0(p, p, 0). The paper only states 'sufficiently regular initial density matrix' and gives no quantitative bound on the corrections. For a generic normalizable state, the error from (ii) is O(q) = O(1/log t) relative, which vanishes, but this is not shown uniformly in p, and for states whose momentum-space coherences vary on a scale ~1/log(rt) the correction could be of the same order as the leading term. Moreover, the abstract's claim that the scale log(rt)/r is 'independent of the initial state' is stronger than what Appendix C actually proves, since the derivation itself assumes regularity. This does not invalidate the central solution, but it means the universal gapless prediction is not fully rigorous as stated.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper introduces a quantum version of stochastic resetting with uniform memory, in which each reset returns the system to a state selected uniformly from its entire past history. The authors formulate an integro-differential master equation (Eq. (5)) and solve it exactly in the energy eigenbasis: every density-matrix element evolves as ρ_{r,mn}(t)=ρ_{r,mn}(0) f(ω_{mn},t), where f is a Kummer function (Eq. (13)). They then separate the long-time behavior into gapped systems, where nonzero Bohr frequencies are bounded below by a gap and coherences decay algebraically while populations are frozen, and gapless systems, where they claim the position distribution spreads universally as log(rt)/r. Illustrative examples include a two-level system, a harmonic oscillator, and a free particle, with Monte Carlo verification for the two-level polarization.","tokens_in":23421,"tokens_out":8513,"duration_ms":73176,"significance":"If correct, this is a valuable contribution: it provides a rare exact solution of a genuinely non-Markovian, nonunitary quantum dynamics. The reduction of the entire problem to a scalar Kummer equation is elegant, and the spectral classification into gapped/gapless regimes is physically illuminating. The gapped-sector results are well supported by the exact solution and the numerical simulation. The paper also makes useful connections to the classical preferential-relocation literature. The main weakness is the derivation of the universal gapless spreading, which is not sufficiently rigorous and currently overstates its independence from the initial state.","major_comments":[{"comment":"The replacement ρ0(p+q/2, p−q/2, 0) → ρ0(p, p, 0) is not valid in general. For a pure state with a spatial displacement, ψ(p)=e^{-ipx0}φ(p), the off-diagonal matrix element contains a phase e^{-iqx0}. For q ~ 1/log(rt) and x0 of order log(rt) or larger, this phase is not negligible, so the approximation drops a factor that shifts the final position distribution by x0. The resulting Eq. (49) is therefore not translation-covariant, and the claim that the position distribution spreads 'independently of the initial state' is not supported. Please either include such phase factors systematically or restate the result as a statement about the shape relative to the initial centroid, with explicit assumptions on the initial state.","section":"Appendix C, Eqs. (C.5)–(C.10)"},{"comment":"The derivation of Eq. (49) rests on an uncontrolled joint limit: t→∞ with ω log(rt) fixed, plus the small-q expansion of the dispersion and the replacement of off-diagonal initial coherences by diagonal ones. The text only says 'sufficiently regular initial density matrix' and gives no explicit error bounds. For a generic normalizable state, the off-diagonal term can vary on a scale comparable to 1/log(rt), in which case the correction is of the same order as the leading term. Please state precise regularity assumptions and provide quantitative estimates for the corrections to Eq. (C.10), uniformly in p. Without this, the universal log(rt)/r scaling is not fully established.","section":"Appendix C, general joint limit"},{"comment":"The abstract and conclusions describe gapped systems simply as 'systems with a discrete energy spectrum'. However, the actual condition used in the proof is the existence of a finite spectral gap Δ>0 for all nonzero Bohr frequencies (Eq. (20)). A discrete spectrum can accumulate (e.g., at a threshold), giving arbitrarily small nonzero frequencies and hence no gap. Please align the wording with the mathematical hypothesis, or explicitly discuss the accumulation-point case.","section":"Abstract and Sec. 5"}],"minor_comments":[{"comment":"Typo: 'differnce' should be 'difference'.","section":"Page 10, Sec. 2, dephasing paragraph"},{"comment":"The phrase 'a the “Master equation”' contains a redundant article; should be 'the “Master equation”'.","section":"Appendix A title"},{"comment":"The expansion ε(p+q/2)−ε(p−q/2)≈q v(p) should explicitly mention the remainder term and the order of the error, especially for power-law dispersions with 1<ν<2 where the function is only C^1 at p=0.","section":"Eq. (C.6)"}],"recommendation":"major_revision","confidential_remarks":"The exact solution (13) and the gapped-sector analysis are solid and form the core value of the paper. The gapless scaling result is interesting but currently overreaches; it needs either a rigorous error estimate with explicit regularity assumptions or a clearly softened statement of universality. I would support acceptance after such a revision."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"First, the load-bearing result is Eq. (13): every density-matrix element in the energy basis evolves as rho_mn(0) M(i omega/(r+i omega); 1; -(r+i omega)t). The derivation from the integro-differential equation is clean, the Kummer identification checks out, and the gapped-sector consequences follow naturally: populations stay frozen, coherences decay algebraically with exponent omega^2/(r^2+omega^2) and oscillate in log t, and the stationary state is the diagonal part of the initial density matrix, independent of r. The two-level Monte Carlo in Fig. 1 supports the exact formula. That part is done carefully and is new—previous quantum resetting work was memoryless, and the classical preferential relocation papers did not handle density-matrix coherences. The citation list is heavy on the authors' own group's earlier work, but those are the relevant papers and the overlap is handled credibly.\n\nThe gapless section is the soft spot. The scaling form (49) is derived in Appendix C by taking a joint limit t→∞, omega→0 with omega log(rt) fixed. The two approximations—replacing the Kummer function by exp(-i omega log(rt)/r) and replacing rho0(p+q/2, p-q/2, 0) by rho0(p,p,0)—are standard, but the paper gives no uniform error bound. 'Sufficiently regular initial density matrix' is doing a lot of work. For momentum-space coherences varying on a scale ~1/log(rt), the correction could contribute at the same order as the leading term. So the universal log(rt)/r scale is plausible and likely right, but the claim that it holds 'independently of the initial state' is stronger than what the appendix proves. This is a request for an explicit estimate or a counterexample, not a reason to reject.\n\nTwo smaller presentation issues. The abstract maps gapped systems to discrete spectra and gapless to continuous spectra; the operative condition is Eq. (20), a positive minimum Bohr frequency. A discrete spectrum can fail that, so the abstract's wording is imprecise. Also, the abstract's 'independent of the initial state' refers only to the scale; the shape of the spreading distribution depends on the initial momentum distribution via V0. The main text is clear about this, the abstract is not.\n\nWho is this for: people in quantum resetting, open quantum systems, and non-Markovian dynamics. It deserves a serious referee. The central calculation is self-contained, new, and supported by numerics. I would send it out and ask for the Appendix C estimate plus abstract revisions.","headline":"Exact solution for quantum resetting with uniform memory is solid and new, but the claimed universal log(rt)/r spreading rests on a joint limit whose error terms are not pinned down; worth refereeing as is.","tokens_in":23867,"tokens_out":2608,"would_cite":true,"duration_ms":22664,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"This paper solves quantum resetting to a random past state exactly: the Hamiltonian enters only through Bohr frequencies, and the long-time fate is dictated by whether the spectrum is discrete (gapped) or continuous (gapless).","keywords":["quantum resetting","memory kernel","non-Markovian dynamics","Kummer confluent hypergeometric function","Bohr frequencies","gapped and gapless spectra","ultraslow spreading","preferential relocation model"],"falsifier":"Run the paper's own event-driven Monte Carlo for a two-level system with Ω=1, r=0.5, starting from |↑_z⟩, and check at very long times that the coherence follows |ρ_{↑↓}(t)| ∝ t^{-4Ω²/(r²+4Ω²)} with oscillations periodic in log t; if instead the coherence decays exponentially, reaches a nonzero plateau, or the diagonal elements change, the exact solution Eq. (13) and the gapped/gapless classification would be refuted.","tokens_in":23008,"feed_emoji":"⚛️","tokens_out":14085,"duration_ms":100965,"temperature":0.7,"pith_summary":"The paper introduces a quantum resetting protocol in which each reset returns the system to a state it occupied at a uniformly chosen past time, making the dynamics nonunitary and non-Markovian. It proves an exact formula for every density-matrix element in the energy basis: the Hamiltonian appears only through the Bohr frequencies, and each element factorizes as its initial value times a single universal function. From this factorization follows a sharp dichotomy: in gapped systems, coherences decay algebraically and the state relaxes to the energy-diagonal part of the initial density matrix, independent of the resetting rate; in gapless systems, the spatial distribution spreads indefinitely on the ultra-slow scale log(rt)/r. In both cases the state retains a strong memory of its initial condition, unlike conventional resetting. A sympathetic reader cares because the framework is fully general for time-independent Hamiltonians and provides a controlled route to ultraslow, memory-preserving quantum dynamics.","feed_headline":"Memory resetting slows quantum dephasing to a power law","feed_subtitle":"Gapped systems relax to initial energy populations; gapless ones spread as log(rt)/r — memory survives.","key_machinery":"The key object is the single scalar function f(ω,t)=M(iω/(r+iω);1;-(r+iω)t), which solves the integro-differential equation for one matrix element after the memory kernel is eliminated by differentiation. Every density-matrix element is ρ_{r,mn}(t)=ρ_{r,mn}(0) f(ω_{mn},t), so all many-body or few-body complexity reduces to evaluating this one function at the Bohr frequencies. The classification follows from its asymptotics: for fixed nonzero ω, |f| decays as t^{-ω²/(r²+ω²)} with log-periodic phase; in the joint limit ω→0 with ω log(rt) fixed, f≈e^{-iω log(rt)/r}, which produces the universal logarithmic spreading in the gapless case.","core_discovery":"The central result is Eq. (13): in the energy eigenbasis, the reset-averaged density matrix evolves as ρ_{r,mn}(t)=ρ_{r,mn}(0) M(iω_{mn}/(r+iω_{mn}); 1; -(r+iω_{mn})t), where M is Kummer's confluent hypergeometric function and ω_{mn}=E_m-E_n are the Bohr frequencies. The Hamiltonian enters only through these energy differences. For a discrete nondegenerate spectrum with minimal gap Δ, populations stay fixed while coherences decay algebraically with exponent ω²/(r²+ω²) and log-periodic oscillations; the stationary state is the infinite-time average of the reset-free unitary dynamics, equal to the initial energy populations and independent of rate r. For a continuous spectrum, arbitrarily smal","pith_inferences":["Because the gapped stationary state is the dephased diagonal ensemble reached without an external bath, this protocol could serve as a built-in dephasing engine for quantum information or precision-measurement settings — a use the paper does not discuss.","The universal log(rt)/r spreading in gapless systems is slow enough to be cleanly separated from ballistic or diffusive spreading over a wide time window, suggesting a direct experimental test in cold atoms or trapped ions where resets to past states can be implemented via stored copies.","For nonuniform memory kernels (e.g., biased toward recent or distant pasts, as the paper suggests), the special-function solution should generalise, but the gapped/gapless dichotomy — algebraic relaxation versus persistent logarithmic spreading — is likely to survive as long as the kernel is normalized and the spectral criterion still holds.","The factorization suggests a single-particle reduction: for Hamiltonians whose spectrum is known (free fermions, integrable chains), the same framework applies with Bohr frequencies given by differences of single-particle energies, so memory resetting may be tractable in many-body settings beyond the examples given."],"forward_implications":["In any gapped system, the stationary state is the initial energy populations (diagonal of the initial density matrix), independent of the resetting rate r, so the protocol acts as a controlled dephasing that preserves all population information.","Relaxation to that state is algebraic, not exponential, with exponent Δ²/(r²+Δ²) set by the smallest Bohr frequency gap, and it is accompanied by oscillations periodic in log t.","In gapless systems, the position distribution never settles; it spreads with the universal scale log(rt)/r, which does not depend on the Hamiltonian or the initial state, although the asymptotic shape of the distribution does.","The factorization ρ_{r,mn}(t)=ρ_{r,mn}(0)f(ω_{mn},t) means any observable in a gapped system can be computed by summing the initial density-matrix elements times one special function, bypassing the full dynamics.","The gapped stationary state coincides with the infinite-time average of the reset-free unitary dynamics, so the memory protocol reproduces time-averaged (dephased) ensembles without any environmental coupling."],"fun_headline_variants":["Quantum memory resets: coherences decay algebraically in gapped systems","Gapless quantum systems spread ultra-slowly under memory resetting","Quantum resetting with memory: Hamiltonian enters via Bohr frequencies only","Populations survive memory resets; coherences decay as power laws","Memory resets turn quantum dephasing algebraic for gapped spectra"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The universal log(rt)/r spreading in gapless systems relies on the assumption that the initial density matrix is 'sufficiently regular' so that off-diagonal coherences and higher-order momentum corrections do not contribute on the same scale as the leading term in the joint limit ω log(rt) fixed — a condition the paper states but does not quantify or bound.","fun_headline_variants_meta":{"raw":{"variants":["Quantum memory resets: coherences decay algebraically in gapped systems","Gapless quantum systems spread ultra-slowly under memory resetting","Quantum resetting with memory: Hamiltonian enters via Bohr frequencies only","Populations survive memory resets; coherences decay as power laws","Memory resets turn quantum dephasing algebraic for gapped spectra"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.00042,"raw_usage":{"total_tokens":2032,"prompt_tokens":816,"completion_tokens":1216,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":560,"completion_tokens_details":{"reasoning_tokens":1123}},"tokens_in":560,"tokens_out":1216,"duration_ms":9906,"temperature":1.0,"reasoning_tokens":1123,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-04T09:46:53.654315+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Run the paper's own event-driven Monte Carlo for a two-level system with Ω=1, r=0.5, starting from |↑_z⟩, and check at very long times that the coherence follows |ρ_{↑↓}(t)| ∝ t^{-4Ω²/(r²+4Ω²)} with oscillations periodic in log t; if instead the coherence decays exponentially, reaches a nonzero plateau, or the diagonal elements change, the exact solution Eq. (13) and the gapped/gapless classification would be refuted.","supporting_citations":[],"review_version":1}