{"id":"8649adc8-6381-4d0b-bf11-dd4dd9aabbdf","arxiv_id":"2607.28730","paper_version":1,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":4,"one_line_summary":"In U(1) and SU(2) lattice gauge theories in the disorder-free localization regime, stabilizer and participation entropies grow as log log t at strong coupling, bounded by an exact counting of resonant configurations.","lead":"This paper tracks how quantum 'magic' and wavefunction spread grow over time in 1+1D lattice gauge theories, finding an extremely slow double-logarithmic growth at large gauge coupling. It matters because the growth rate of these resources determines when quantum simulation of gauge theories becomes hard to mimic classically.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Double-log 'growth' is supported only by model selection; the configuration-space argument is an upper bound and supplies no lower bound, so the unbounded law is not substantiated.","rationale":"The paper's central novelty is the claimed ultraslow double-logarithmic growth of SRE2/PE2 at strong coupling. The reader's conditional verdict already emphasizes that the bound does not prove the growth law and that the classification is statistical over finite windows. My stress-test sharpens this: even granting the paper's Δε=0 shell and the logarithmic perturbative depth, the analytical result is only an upper envelope. A lower bound—or some control on the actual occupation of the resonant configurations—would be required to substantiate unbounded growth. The numerical evidence (P≈0.90–0.95) is suggestive but not decisive, especially given the SM's explicit caveat that the diagnostics do not establish the asymptotic law and that no endpoint-shift stability analysis is provided. This is not an internal inconsistency; the bound and exact counting are independently valuable. But the headline claim is stronger than what the analysis supports. Because the reader already conditions the verdict on essentially this issue, the verdict should remain CONDITIONAL rather than being upgraded or rejected.","tokens_in":39156,"tokens_out":5629,"duration_ms":65791,"concrete_test":"Refit the U(1) SRE2 and PE2 traces at g²=16,20 using the four candidate laws (S11a)–(S11d) on nested windows [10²,10^M] with M=13,14,15, recomputing the bootstrap selection probability and cross-validated slope for each M. If the double-log model's P drops below ~0.7, or if the fitted slope a changes by more than 50% as M is varied, the numerical double-log classification is a finite-window effect and the claim should be weakened from 'growth' to 'compatible with an upper ceiling.'","verdict_should_be":"UNCHANGED","load_bearing_attack":"The most load-bearing weakness is that the analytic configuration-space argument is a one-sided upper bound, not a derivation of growth. Equations (S26), (S42), and (S43) give PE2(t) ≤ log V_res(n*(t)) ~ k log log t + const and hence SRE2 ≤ 2 PE2 ≤ 2k log log t, but these are ceilings: a state that remains essentially frozen, or one whose support is exponentially smaller than V_res, satisfies them equally well. The exact transfer-matrix count establishes only that the number of Δε=0 configurations within perturbative distance n*(t) grows as r^k; it says nothing about how much probability actually populates those configurations. Consequently, the qualitative claim that SRE2/PE2 grow as a log log t + b rests entirely on the bootstrap classification over finite windows (Table S2). The authors themselves state that these diagnostics 'do not establish the t→∞ asymptotic law' and no window-shift stability analysis is reported, so the headline 'double-logarithmic growth, substantiated by a configuration-space bound' is overclaimed. The bound is real and useful as a ceiling, but it cannot distinguish unbounded growth from saturation below the ceiling.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies the real-time growth of three quantum-resource measures — stabilizer Rényi-2 entropy, participation Rényi-2 entropy, and fermionic non-Gaussianity — in 1+1D U(1) and truncated SU(2) lattice gauge theories in the disorder-free localization regime. Exact time evolution is performed for U(1) up to t=10^15 at L=16 and for SU(2) up to t=10^4 at L=12, with several hundred superselection sectors averaged. A statistical model-selection protocol (local logarithmic slopes, forward cross-validation, block bootstrap, AIC) identifies saturating power-law relaxation at intermediate couplings and favors unbounded double-logarithmic growth for SRE2 and PE2 at the largest U(1) couplings and for PE2 at strong SU(2) couplings, while NG saturates everywhere. The double-logarithmic law is connected to a configuration-space argument: strong-coupling perturbation theory gives a logarithmic perturbative depth n*(t), and an exact transfer-matrix count of resonant configurations gives a polynomial resonant volume Vres(r) ~ r^k with k ≈ 0.145L, yielding PE2 ≤ k log log t and SRE2 ≤ 2 PE2. The paper concludes that gauge invariance constrains the generation of quantum resources.","tokens_in":39423,"tokens_out":4547,"duration_ms":52344,"significance":"If the double-logarithmic growth claim is correct, this is a valuable addition to the DFL literature and to the quantum-resource dynamics of gauge theories, with concrete implications for classical simulability and for near-term quantum simulators. The paper has genuine strengths: the exact transfer-matrix counting of resonant configurations up to L=96 is an impressive, reproducible piece of combinatorics; the inequality SRE2 ≤ 2 PE2 is proven exactly for arbitrary pure states; and the statistical analysis is unusually careful, with bootstrap and cross-validation used and the finite-window limitation explicitly acknowledged. These strengths make the numerical study a solid contribution. However, as detailed below, the configuration-space bound is only an upper bound and does not, by itself, substantiate unbounded growth; the paper's headline claim is therefore stronger than the evidence supports.","major_comments":[{"comment":"The configuration-space argument provides only a one-sided bound: PE2(t) ≤ log Vres(n*(t)) ~ k log log t and SRE2 ≤ 2 PE2. Every state whose support remains inside a small subset of Vres — including an essentially frozen state — satisfies the same inequality. The bound therefore cannot 'substantiate' unbounded double-logarithmic growth; it only shows that such growth is compatible with the counting. The abstract and the Conclusions ('this law ... substantiated by a configuration-space bound') overstate what is derived. Please either rephrase the claim as 'an ultraslow ceiling consistent with the observed growth' or add a lower bound / typicality argument showing that the wave function actually populates a non-vanishing fraction of the resonant volume.","section":"SM, Eqs. (S26), (S42)-(S44)"},{"comment":"The double-logarithmic classification at U(1) g²=16,20 rests on bootstrap probabilities P≈0.90-0.95 evaluated on a single fixed time window [10²,10^15]. The SM itself states that these diagnostics 'do not establish the t→∞ asymptotic law' and reports no window-shift stability analysis. Because the central claim of the Letter is the unbounded asymptotic law, this is load-bearing. Either add a stability analysis under shifts of t_min and t_max (e.g., showing the selected model and P remain stable over a range of endpoints), or consistently label the conclusion as 'slow growth within the simulated window' in the abstract and main text, not as an established asymptotic law.","section":"Table S2 and SM 'Model-selection results'"},{"comment":"The derivation of n*(t) ~ log(wt)/log(g²/w) assumes that inter-tower transitions are purely virtual and that only configurations with zero electric-energy mismatch (Δε=0) are permanently accessible. This is the key assumption that makes the bound an upper bound; if multi-hop resonances or finite-size leakage populate off-resonant configurations, the ceiling collapses. No numerical check of this assumption is reported. Please quantify the total probability weight on configurations with Δε≠0 as a function of g² and time in the exact time evolutions of Fig. 2, or otherwise justify that the leakage is negligible on the simulated window.","section":"SM, 'Strong-coupling dynamics and the logarithmic perturbative depth'"},{"comment":"In the SU(2) theory the double-logarithmic classification is much weaker: the window spans only ~3 decades and the bootstrap probabilities at the largest couplings are P=0.59-0.73, with P=0.95 only at g²=7.75. The main text says the analysis 'favors continued double-logarithmic growth' at g²≥5.33, but the Conclusion states the ultraslow growth holds 'for both gauge groups'. This generalization goes beyond the reported evidence. Please temper the SU(2) claim to reflect the indicative, not conclusive, nature of the classification, as already acknowledged in the text.","section":"Main text, 'The SU(2) dynamics' / Table S2"}],"minor_comments":[{"comment":"The caption says data and fit curves are displayed up to t=10^11, whereas the fit parameters are determined on [10²,10^15]. Please clarify whether the simulated data extend to 10^15 and are only shown up to 10^11, or whether the fits use data not displayed.","section":"Fig. 2 caption"},{"comment":"The PE2 relaxation-exponent trend has R²_adj=0.737 over five couplings; the text already cautions against over-interpretation. It would be helpful to state the fit window and the number of points in the caption or table note.","section":"End Matter, Table I"},{"comment":"The effective sample size uses the lag-1 autocorrelation c_1 of the residuals, but the formula reduces N by a factor that depends on c_1. Since AIC is used only as a cross-check, consider stating that the qualitative conclusions are unchanged under Neff/N in [0.2,1] to make the robustness explicit.","section":"SM, Eq. (S21)"}],"recommendation":"major_revision","confidential_remarks":"This is a borderline case. The numerical work is careful and the exact counting is a real contribution, but the central claim — 'ultraslow double-logarithmic growth substantiated by a configuration-space bound' — is overstated, as the bound is only an upper bound and the asymptotic law is not established. The manuscript would be greatly strengthened by an endpoint-shift stability analysis, a quantitative check of the Δε=0 assumption, and a consistent rewording of the abstract and Conclusions that separates 'observed slow growth in the window' from 'proven unbounded growth'. As it stands, I cannot recommend acceptance without these changes; the issues are fixable within the scope of the paper, hence major revision rather than rejection."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The short version: this is a careful numerical paper with two clean analytical byproducts, and the central physical claim—ultraslow double-log growth of stabilizer and participation entropies at strong coupling—is plausible but not actually proven. If you read it as \"we observe a very slow growth consistent with log log t and we give a rigorous ceiling for how fast it could ever grow,\" it holds up. If you read the abstract's \"substantiated by a configuration-space bound\" as a derivation of the growth law, it oversells.\n\nWhat is genuinely new: the first study of stabilizer Rényi entropy, participation entropy, and fermionic non-Gaussianity in disorder-free localization of U(1) and SU(2) gauge theories; an exact inequality SRE2 ≤ 2 PE2 that follows from the Pauli-weight marginal identity; and an exact transfer-matrix counting of the resonant volume that gives a polynomial growth with exponent k ≃ 0.145L. The exact counting is the best part of the paper—it is reproducible, verified against brute force for small L, and gives a real, independent constraint on how many configurations are even accessible. The statistical analysis in the supplement is also honest: bootstrap, cross-validation, local slopes, and they state plainly that their diagnostics do not fix the t→∞ law.\n\nSoft spots, in proportion. The configuration-space argument is a one-sided bound. It counts zero-mismatch configurations within perturbative depth n*(t), but it does not show how much probability actually populates those configurations. A state that stays nearly frozen satisfies the bound equally well. So the double-log law itself rests on the bootstrap classification over finite windows (P ≈ 0.9–0.95 for the two largest U(1) couplings), not on the counting. The authors do flag this in the supplement, but the abstract and main text lean harder on the bound than it can bear. The SU(2) results are explicitly indicative, with only about three decades, and no code or data with a commit hash is provided, which matters for a paper whose main evidence is numerical. The static finite-size crossing is a nice consistency check but is itself limited to L ≤ 12.\n\nNone of these are fatal. The exact inequality and the exact counting are results I would want to cite independently, and the numerical work is careful enough to deserve referee time. The right outcome is acceptance after revision that softens the claim and provides the data.","headline":"Solid numerical study with one genuinely useful exact inequality, but the headline double-log growth is only an upper-bound ceiling and not a proven asymptotic law.","tokens_in":39933,"tokens_out":1086,"would_cite":true,"duration_ms":14943,"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":"Gauge invariance alone can slow the growth of quantum complexity to a double-logarithmic rate.","keywords":["disorder-free localization","lattice gauge theory","stabilizer Rényi entropy","participation entropy","fermionic non-Gaussianity","double-logarithmic growth","confinement","quantum resources"],"falsifier":"Measure or compute PE2 at strong coupling (g² ~ 16–20) in a U(1) LGT for times beyond 10^15 and system sizes beyond L=16: if the growth departs from a log log t fit and saturates, or if the population of configurations with nonzero electric-energy mismatch exceeds the perturbative (w/g²)^n bound, the resonant-volume argument fails. Equivalently, directly compute matrix elements connecting different energy towers and check whether they are suppressed only by powers of w/g².","tokens_in":39036,"feed_emoji":"⏳","tokens_out":4727,"duration_ms":43403,"temperature":0.7,"pith_summary":"This paper tries to establish that in (1+1)-dimensional U(1) and SU(2) lattice gauge theories, the constraints of gauge invariance—without any quenched disorder—can slow the growth of quantum complexity to an ultraslow, double-logarithmic law. Simulating quenches from superpositions of superselection sectors (the disorder-free localization protocol), the authors find two regimes: at intermediate gauge coupling, stabilizer and participation Rényi-2 entropies relax to saturation as power laws; at the largest couplings, they continue to grow, but no faster than log log t, while fermionic non-Gaussianity saturates throughout. The double-logarithmic ceiling is derived from an exact counting of resonant configurations: configurations with zero electric-energy mismatch within radius r grow polynomially, V_res(r) ~ r^k with k ≈ 0.145L, and the accessible depth grows only logarithmically in time. The result matters because it identifies gauge symmetry as an intrinsic generator of non-ergodic, classically compressible dynamics.","feed_headline":"Gauge forces cap complexity growth at log log t","feed_subtitle":"Stabilizer and participation entropies slow to a double-logarithmic crawl, tying confinement to classical simulability.","key_machinery":"The central object is the resonant configuration volume V_res(r): the number of computational-basis configurations reachable from the initial configuration whose electric-energy mismatch with the initial state is exactly zero and whose graph distance is at most r. It is evaluated exactly by a transfer-matrix scan over electric-field profiles, and its polynomial growth V_res(r)∼r^k, with k≃0.145L, combined with the logarithmic perturbative depth n*(t)∼log(wt)/log(g²/w), produces the double-logarithmic ceilings PE2≲k log log t and SRE2≤2PE2.","core_discovery":"The authors show that in the strong-coupling regime of both U(1) and SU(2) lattice gauge theories, the stabilizer Rényi-2 entropy and the participation Rényi-2 entropy are bounded by k log log t + const, with k ≃ 0.145L extracted from exact transfer-matrix counting of resonant configurations up to L=96. The argument combines three steps: exact kinematic bounds (PE2 ≤ log|Ω|, SRE2 ≤ log|ΔΩ|, and SRE2 ≤ 2PE2), degenerate perturbation theory giving a logarithmic perturbative depth n*(t) ~ log(wt)/log(g²/w), and the geometric finding that the resonant volume grows polynomially with exponent k, corresponding to a finite density of mobile charges. Fermionic non-Gaussianity, by contrast, relaxes to","pith_inferences":["If the bound holds in the thermodynamic limit, the density of complexity cost—PE2/L and SRE2/L—vanishes logarithmically with time, suggesting these gauge-theory quenches become progressively easier to simulate classically as the system grows.","The finite density of mobile charges (k ≈ 0.145 per site) may be a diagnostic of the confinement length scale; measuring k from the slope of PE2 versus log log t would provide a new dynamical probe of confinement.","The argument implies that any mechanism that violates the zero-mismatch restriction—such as multi-phonon resonances or string-breaking processes—would convert the double-log law into a single log or a power law, so the ultraslow growth is a sharp test of confinement.","One could extend the resonant-volume counting to higher dimensions or to Z2 gauge theories to see whether the k∝L extensivity and the double-log ceiling survive beyond one spatial dimension."],"forward_implications":["At strong coupling, stabilizer and participation Rényi-2 entropies grow no faster than a double logarithm in time, so the Pauli-weight distribution remains confined to a polynomially growing set of strings.","The same ultraslow law appears in both Abelian U(1) and non-Abelian SU(2) theories, making it a gauge-invariance-driven effect rather than a feature of a specific gauge group.","Fermionic non-Gaussianity saturates at every coupling, so two-point correlations remain near-Gaussian and the dynamics may be amenable to Majorana-propagation simulation even where wavefunction spreading is slow.","The double-logarithmic curves shift downward by k log log(g²/w) as the coupling grows, giving a quantitative, testable prediction for how the slowdown depends on coupling.","All three complexity measures are measurable in current quantum simulators, so the ultraslow growth is an experimental target rather than a purely numerical observation."],"fun_headline_variants":["Strong gauge coupling slows complexity to log log t","Gauge theories cap quantum complexity growth at log log t","Double-log complexity growth in lattice gauge theories","Gauge invariance constrains complexity to ultra-slow growth","Quantum complexity crawls in strong-coupling gauge dynamics"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The double-logarithmic ceiling relies on the assumption that, at strong coupling, the state can permanently access only configurations whose electric energy exactly matches the initial configuration, with every transition between different energy towers remaining virtual; if multi-hop resonances allow real leakage between towers, the ceiling collapses.","fun_headline_variants_meta":{"raw":{"variants":["Strong gauge coupling slows complexity to log log t","Gauge theories cap quantum complexity growth at log log t","Double-log complexity growth in lattice gauge theories","Gauge invariance constrains complexity to ultra-slow growth","Quantum complexity crawls in strong-coupling gauge dynamics"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000162,"raw_usage":{"total_tokens":1074,"prompt_tokens":739,"completion_tokens":335,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":483,"completion_tokens_details":{"reasoning_tokens":260}},"tokens_in":483,"tokens_out":335,"duration_ms":3559,"temperature":1.0,"reasoning_tokens":260,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-03T00:31:51.686355+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure or compute PE2 at strong coupling (g² ~ 16–20) in a U(1) LGT for times beyond 10^15 and system sizes beyond L=16: if the growth departs from a log log t fit and saturates, or if the population of configurations with nonzero electric-energy mismatch exceeds the perturbative (w/g²)^n bound, the resonant-volume argument fails. Equivalently, directly compute matrix elements connecting different energy towers and check whether they are suppressed only by powers of w/g².","supporting_citations":[],"review_version":1}