{"id":"5be695e5-f128-409b-a0ba-910aee0eb62c","arxiv_id":"2412.19801","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":5,"one_line_summary":"Ergotropy of large many-body quantum batteries concentrates exponentially around its average for almost all random states, with numerical evidence for the same behavior under the Bures (least-informative) prior.","lead":"For large quantum batteries, the amount of extractable work (ergotropy) is almost the same for almost every random quantum state, so a typical state stores a macroscopic charge. The result, proven for broad many-body systems and supported by numerics for a least-informative prior, suggests quantum battery charge may be noise-robust without fine-tuning.","discovery_kind":"first_principles","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Macroscopic-ergotropy claim rests on unproven Eq. (39), which fails for local Hamiltonians with a narrow normalized spectrum such as the number operator; the concentration theorem itself survives.","rationale":"The reader's conditional verdict correctly identifies Eq. (39) as the weakest point. My stress-test strengthens this: the missing hypothesis is not just unproven but violated by standard local Hamiltonians with a narrow normalized density of states. The number-operator example satisfies every assumption of Sec. I and Eq. (38), yet its normalized spectrum has width O(1/sqrt N), making the average ergotropy vanish as N grows. Therefore the paper's statement that macroscopic ergotropy holds for 'very general and inclusively defined local Hamiltonians' is false without a non-degeneracy condition. The core concentration theorem (Theorem 1 plus the Levy lemma application plus Eq. (2) scaling) remains valid: the Lipschitz proof is correct, the purification map preserves the Hilbert-Schmidt measure, and the bound Upsilon_E = O(N^k/e^{kappa N}) follows. The overclaim is localized to Sec. V and the introduction's 'macroscopic' sentence; the abstract's concentration claim is safe. Thus a conditional acceptance requiring revision is appropriate: either add an explicit assumption that the normalized spectral density has O(1) width, or reframe macroscopic ergotropy as a property of Hamiltonians with many distinct eigenvalues and demote the nGUE numerics to supporting evidence for that narrower class. The Bures-measure section's 'we have therefore shown' also needs softening to 'numerical evidence indicates,' as the reader noted.","tokens_in":17788,"tokens_out":19069,"duration_ms":197942,"concrete_test":"Compute <E_hat>_HS for the N-qubit number operator H_hat = (1/N) sum_i |1><1|_i by sampling HS states via Eq. (40) for N = 10 to 20 (d up to about 10^6), using at least 10^5 samples for small N and 10^4 for large N. Plot <E_hat>_HS versus N alongside the nGUE curve from Fig. 1. If the number-operator average decays as N^{-1/2} or vanishes while the nGUE average stays above 0.23, Eq. (39) cannot hold for all allowed local Hamiltonians, and the macroscopic claim must be restricted. An analytic cross-check is to combine the Marchenko-Pastur law for HS eigenvalues with the binomial density of states to show the passive energy equals 1/2 - c/(2 sqrt N).","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The analytic concentration bound (Eqs. (24) and (35)) is sound; the problem is the macroscopic-mean claim. Eq. (39) -- existence of epsilon with <E_hat>_HS >= epsilon for all d -- is used to conclude <E>_HS is proportional to ||H||_op (Eq. (42)). This is not merely unproven: it is false for local Hamiltonians allowed by the stated assumptions. Take H = sum_{i=1}^N |1><1|_i on N qubits (k=1, ||H||_op = N, ln d = N ln 2, and Eq. (38) holds with h = 1/2). The normalized spectrum H/N has eigenvalues m/N with degeneracy C(N,m), i.e. a Gaussian density of states of width O(1/sqrt N) around 1/2. For HS-random rho, the passive energy E[Tr(rho^down H_hat)] tends to the average energy Tr(H_hat)/d = 1/2 because the normalized energies coalesce; hence <E_hat>_HS ~ c/sqrt N -> 0, so Eq. (39) fails. This is not an exotic case: it is a standard 1-body spin Hamiltonian. Footnote 53 excludes it by assuming the number of distinct eigensubspaces is proportional to d, but Secs. I and V claim the macroscopic result for all k-body Hamiltonians with bounded terms, and no proof of Eq. (39) under that or any other stated condition is given. The numerical support uses nGUE, whose normalized spectrum has O(1) width, not the O(1/sqrt N) width typical of local Hamiltonians. The concentration theorem survives; the macroscopic-ergotropy conclusion does not, unless an explicit spectral non-degeneracy assumption is added.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper proves that ergotropy, as a function of the quantum state, is Lipschitz continuous with respect to trace, Bures, and Hilbert-Schmidt distances (Theorem 1, Appendix A). Using the purification correspondence between Hilbert-Schmidt-random states and Haar-random pure states on a larger space, the authors apply Levy's lemma to show that for large Hilbert-space dimension d the ergotropy of a state drawn from the Hilbert-Schmidt measure is exponentially concentrated around its ensemble average (Eqs. (24), (25), (29), (30)). For k-body Hamiltonians with ln d proportional to the particle number N, the absolute width of the ergotropy distribution is O(N^k/e^{kappa N}), so the concentration is exponentially strong in N (Eq. (35)). The paper further argues that the average ergotropy is macroscopic, proportional to ||H||_op (Eq. (42)), and provides numerical evidence for analogous concentration when states are sampled from the Bures measure (Section VII).","tokens_in":18226,"tokens_out":8112,"duration_ms":88384,"significance":"If the macroscopic-average claim were established under the stated assumptions, the paper would be a significant contribution to quantum thermodynamics: it would show that typical highly mixed states of many-body batteries contain a macroscopic amount of unitarily extractable work and that this amount is stable under small perturbations. The technical core of the paper, Theorem 1 and the concentration argument in Section IV, is sound and clearly presented; the Lipschitz proof via Lidskii-Wielandt and Fuchs-van de Graaf is clean. The numerical Bures-measure section is honestly labeled as evidence rather than proof. The main weakness is that the macroscopic-ergotropy conclusion relies on an unproved and, as stated, false lower bound on the average normalized ergotropy, so the advertised claim is broader than what is actually demonstrated.","major_comments":[{"comment":"The claim that <E_hat>_HS >= epsilon for all d (Eq. (39)) is load-bearing for the macroscopic-ergotropy conclusion (Eq. (42)) and for the discussion in Section VIII, but it is not proved and it is false for Hamiltonians satisfying the stated assumptions of Section I. Consider H = sum_{i=1}^N |1><1|_i on N qubits, with ||H||_op = N, k=1, and ln d = N ln 2. This Hamiltonian satisfies Eq. (38) with h = 1/2. Its normalized spectrum has eigenvalues m/N with binomial degeneracies, i.e., a Gaussian density of states of width O(N^{-1/2}) centered at 1/2. For a Hilbert-Schmidt-random state rho, the passive energy Tr(rho_down H_hat) tends to the average energy Tr(H_hat)/d = 1/2, so <E_hat>_HS tends to zero as N grows and Eq. (39) fails. Footnote 53 excludes this case by assuming that the number of distinct eigensubspaces is proportional to d, but that assumption is not part of the hypotheses in Section I or of Theorem 1, and no proof of Eq. (39) is given under it. The numerical support uses nGUE Hamiltonians, whose normalized spectrum has O(1) width, and thus does not probe the O(1/sqrt N) spectral width typical of local Hamiltonians. The concentration theorem in Eqs. (24) and (35) survives, but the macroscopic-ergotropy result must either be restricted to Hamiltonians with an explicit spectral non-degeneracy condition or be rephrased as a numerical observation for the ensembles tested.","section":"Sec. V, Eqs. (37)-(42), footnote 53"}],"minor_comments":[{"comment":"Equation (27) drops the 1/ln 2 factor from Eq. (19): with L_E^S <= sqrt(8 ln d)/ln 2, one obtains Upsilon_S <= sqrt(100 pi ln d)/(ln 2 d), not sqrt(100 pi ln d)/d. This is a constant-factor error and does not change the qualitative concentration statement, but the displayed inequality as written is too strong.","section":"Sec. IV, Eq. (27)"},{"comment":"The assumption that the number of distinct eigensubspaces is proportional to d is stated only in a footnote; since it is needed for the macroscopic-average argument, it should appear in the main text as an explicit hypothesis in Section I and in the theorem statements.","section":"Sec. V, footnote 53"},{"comment":"The reference for the lower bound in Eq. (A8) contains unresolved placeholder citations \"[? ? ]\"; these references need to be completed.","section":"Appendix A, after Eq. (A8)"},{"comment":"There is a typo: \"independent, identically identically distributed\" should read \"independent and identically distributed.\"","section":"Sec. V, paragraph on Ginibre matrices"},{"comment":"The word \"noniformativeness\" should be \"noninformativeness.\"","section":"Appendix C"},{"comment":"The statement that Var_{rho,H}(W) concentrates for Hilbert-Schmidt-random rho is asserted without proof or derivation. Since this is a side remark rather than a central claim, it should be either proved briefly or explicitly labeled as a conjecture.","section":"Sec. VI, Eq. (44)"},{"comment":"The notation in Eq. (35) appears to refer to the absolute width Upsilon_E, but it is written as a ratio Upsilon_E/E; please clarify whether the displayed asymptotic is for the absolute width or for the relative fluctuation, since the two differ by a factor ||H||_op when Eq. (42) holds.","section":"Sec. V, Eq. (35)"}],"recommendation":"major_revision","confidential_remarks":"The concentration theorem is sound and the paper is worth publishing after revision, but the macroscopic-ergotropy claim is currently overbroad and is contradicted by an allowed local Hamiltonian. The fix is within scope: add an explicit spectral non-degeneracy assumption, prove Eq. (39) under that assumption if possible, or substantially soften the macroscopic claim and make clear that it is supported only numerically for nGUE-type spectra. The numerical Bures section is clearly labeled as evidence and does not need to become a proof for acceptance, but the authors should present it as such. The missing 1/ln 2 factor in Eq. (27) and the unresolved reference placeholders should also be corrected."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The real result here is Theorem 1 plus the Levy argument: ergotropy is Lipschitz in the Bures, trace, and Hilbert–Schmidt distances, with clean bounds, and therefore concentrates exponentially under the Hilbert–Schmidt measure for k-local Hamiltonians. Appendix A is rigorous and standard—Lidskii–Wielandt, Fuchs–van de Graaf, all accounted for. That is a new and useful contribution, and the paper is honest about what it proves versus what it conjectures, at least in the main text.\n\nWhat I would push back on is the leap from concentration to macroscopic ergotropy. The claim that <E_hat>_HS >= epsilon for all d is Eq. (39), and the paper itself admits it is unproven. The stress-test example is not exotic: H = sum_i |1><1|_i on N qubits has k=1, satisfies Eq. (38), and yet the normalized spectrum has width O(1/sqrt N), so the HS-averaged passive energy tends to the flat average and <E_hat>_HS -> 0. That kills Eq. (39) for a standard one-body spin Hamiltonian. Footnote 53 tries to exclude this by assuming the number of distinct eigensubspaces is proportional to d, but that assumption is not part of the theorem statement and is not proven for the local Hamiltonians the paper claims to cover. So the concentration theorem stands, but the \"most states store macroscopic ergotropy\" conclusion does not, unless an explicit non-degeneracy condition is added and Eq. (39) is actually proved.\n\nAlso worth noting: the Bures-measure section fits a hypothesized tail law to numerics and then says \"we have therefore shown\" exponential concentration. The abstract is more careful, but the text overstates. That is a framing issue, not a fatal one. Minor slips: Eq. (27) drops a 1/ln 2 factor, and there are broken citation placeholders after Eq. (A8). No code or data shipped, though the numerics are simple enough to reproduce.\n\nThe entropy half is a compilation of known results with the Bures Lipschitz bound adapted from Sekatski et al. Fine, but not new.\n\nBottom line: this is a serious paper with a clean new theorem and an overreach in the interpretation. A referee should engage with it, but the macroscopic claim needs either a proof under stated assumptions or an explicit spectral-condition hypothesis. I would send it to review, expecting major revision. If the authors fix the scope of the macroscopic claim, the concentration theorem alone justifies publication.","headline":"Solid new concentration theorem for ergotropy, but the macroscopic-average conclusion leans on an unproven assumption that fails for some perfectly local Hamiltonians.","tokens_in":18737,"tokens_out":1085,"would_cite":true,"duration_ms":13091,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["15B52","60B20","81P45"],"pacs":["03.65.-w","05.30.-d"],"model":"deepseek-v4-flash","headline":"This paper proves that the maximal unitarily extractable work from a large many-body quantum system is exponentially concentrated around its average for almost all states, and that this average is macroscopic.","keywords":["ergotropy","quantum batteries","measure concentration","Levy's lemma","Bures measure","Hilbert-Schmidt measure","von Neumann entropy","many-body systems"],"falsifier":"Compute ⟨Ê⟩_HS for nGUE Hamiltonians at dimensions far beyond d ≈ 8200, for example d = $10^{5}$ or larger, using Hilbert-Schmidt state sampling; if the average normalized ergotropy decreases monotonically to zero rather than saturating at a positive constant, the macroscopic-charge conclusion fails, although concentration itself would survive.","tokens_in":17519,"feed_emoji":"🔋","tokens_out":5679,"duration_ms":60700,"temperature":0.7,"pith_summary":"Ergotropy is the maximal work extractable from a quantum state by cyclic unitary operations, and it is the standard way to quantify the charge of a quantum battery. This paper proves that for large many-body systems whose Hamiltonian has at most k-body interactions, ergotropy concentrates: almost all states sampled from the Hilbert-Schmidt measure have ergotropy within an exponentially small window of the ensemble average. The average is itself macroscopic, scaling with the Hamiltonian's norm rather than decaying with system size, and the same concentration holds for von Neumann entropy. The proof rests on a Lipschitz-continuity estimate for ergotropy in the Bures distance and Levy's concentration lemma. For the Bures measure, the least-informative prior, no analytic bound is available, but the paper's numerics show the same concentration.","feed_headline":"Nearly every state of a large quantum battery has the same ergotropy","feed_subtitle":"Extractable work concentrates exponentially around a macroscopic average in generic many-body systems.","key_machinery":"The load-bearing object is the Lipschitz continuity of ergotropy with respect to the Bures distance, with constant L_E^B ≤ 2||H||_op, proved in Theorem 1 using the Lidskii-Wielandt theorem. Because purification maps Hilbert-Schmidt-sampled states to Haar-uniform points on the sphere $S^{{2d²−1}}$, this continuity combined with Levy's concentration lemma turns Bures-distance control into exponential tail bounds. For entropy, the same route uses the known Lipschitz bound L_S^B ≤ π ln d / ln 2 (equivalently L_E^S ≤ √(8 ln d)/ln 2).","core_discovery":"The paper's central claim is that ergotropy obeys a concentration of measure: for a many-body Hamiltonian with local terms bounded by a constant and at most k-body interactions, and for states drawn from the Hilbert-Schmidt measure, the deviation of ergotropy from its ensemble average decays doubly exponentially. Concretely, Prob(|E(ρ,H) − ⟨E⟩_HS| > ℓ) ≤ 3 $e^{{−ℓ²/Υ_E²}}$ with Υ_E = O(N^k/$e^{{κN}}$) when N ≫ k and ln d ∝ N. The average is not a small residue: numerical evaluation for random nGUE Hamiltonians gives ⟨Ê⟩_HS ≳ 0.23, so ⟨E⟩_HS ∝ ||H||_op, under the macroscopic-energy condition that the infinite-temperature average energy is extensive. The same machinery bounds the concentration of von Neumann entropy, with width Υ_S = O(ln d / d). For the Bures measure, no analytic concentration bound is proven; the paper supplies numerical evidence that both quantities concentrate there as well.","pith_inferences":["If the macroscopic-average condition ⟨Ê⟩_HS ≥ ε fails for some Hamiltonian family, concentration still holds but converges to a vanishing charge; separating which Hamiltonians satisfy this condition is a concrete open problem the paper leaves open.","The result suggests a typicality principle for quantum batteries analogous to canonical typicality for entanglement: for mesoscopic systems that randomize enough, unknown states generically carry near-optimal ergotropy, which may simplify protocols for work extraction from unknown quantum sources.","A testable extension would be to compute the concentration width numerically for structured spin-chain Hamiltonians, such as transverse-field Ising or Heisenberg models, rather than GUE/nGUE spectra, to see how the O(N^k/e^{κN}) bound behaves when k is fixed and N grows."],"forward_implications":["Typical states of a large quantum battery store a macroscopic amount of extractable work; no fine-tuning of the initial state is needed for high charge.","The charge level is noise-robust: small perturbations move the state within the typical set, so close-to-average ergotropy is stable.","The first moment of extracted work is predictable, yet fluctuations of work itself are not suppressed: the noise-to-signal ratio saturates near 1.29.","The same concentration applies numerically to the least-informative Bures measure, suggesting typicality extends beyond Hilbert-Schmidt sampling.","For Hamiltonians satisfying the macroscopic-energy condition, typical ergotropy is proportional to ||H||_op, so the typical battery charge grows with the system's energy scale N^k."],"supporting_citations":[{"why":"Defines ergotropy and the passive state, the object being studied.","marker":"[1]"},{"why":"Supplies the Lidskii-Wielandt theorem used to bound the difference of passive-state eigenvalues in the proof of Theorem 1.","marker":"[43]"},{"why":"Provides the explicit Levy concentration bound with α = 1/(25π) used to obtain the exponential tail estimates.","marker":"[33]"},{"why":"Establishes that Hilbert-Schmidt sampling of density matrices corresponds to Haar-uniform purification vectors, the bridge to Levy's lemma.","marker":"[52]"},{"why":"Gives the Lipschitz bound for von Neumann entropy on the purification sphere used for the entropy concentration result.","marker":"[50]"},{"why":"Supplies the Gaussian Unitary Ensemble construction used to sample random Hamiltonians for the numerical test of macroscopic average ergotropy.","marker":"[55]"},{"why":"Provides the random-matrix construction of Bures-measure states used in the numerical concentration analysis for the least-informative prior.","marker":"[74]"}],"fun_headline_variants":["Ergotropy almost constant for nearly all many-body states","Quantum battery charge concentrates: ergotropy invariance","Most states share same extractable work in large batteries","Ergotropy's measure concentration: a universal law","Large many-body batteries: ergotropy is nearly fixed"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The argument needs the average normalized ergotropy ⟨Ê⟩_HS to stay above some fixed positive value ε for all Hilbert-space dimensions; the paper proves concentration regardless, but without this assumption the concentrated value could be zero, so the battery charge would be trivial.","fun_headline_variants_meta":{"raw":{"variants":["Ergotropy almost constant for nearly all many-body states","Quantum battery charge concentrates: ergotropy invariance","Most states share same extractable work in large batteries","Ergotropy's measure concentration: a universal law","Large many-body batteries: ergotropy is nearly fixed"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000121,"raw_usage":{"total_tokens":1089,"prompt_tokens":938,"completion_tokens":151,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":554,"completion_tokens_details":{"reasoning_tokens":84}},"tokens_in":554,"tokens_out":151,"duration_ms":2709,"temperature":1.0,"reasoning_tokens":84,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-10T23:52:00.429407+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute ⟨Ê⟩_HS for nGUE Hamiltonians at dimensions far beyond d ≈ 8200, for example d = $10^{5}$ or larger, using Hilbert-Schmidt state sampling; if the average normalized ergotropy decreases monotonically to zero rather than saturating at a positive constant, the macroscopic-charge conclusion fails, although concentration itself would survive.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the Lidskii-Wielandt theorem used to bound the difference of passive-state eigenvalues in the proof of Theorem 1."},{"cited_title":"Watrous,The Theory of Quantum Information (Cam- bridge University Press, Cambridge, 2018)","cited_arxiv_id":null,"evidence_quote":"Provides the explicit Levy concentration bound with α = 1/(25π) used to obtain the exponential tail estimates."},{"cited_title":"Życzkowski and H.-J","cited_arxiv_id":null,"evidence_quote":"Establishes that Hilbert-Schmidt sampling of density matrices corresponds to Haar-uniform purification vectors, the bridge to Levy's lemma."},{"cited_title":"Edelman and N","cited_arxiv_id":null,"evidence_quote":"Supplies the Gaussian Unitary Ensemble construction used to sample random Hamiltonians for the numerical test of macroscopic average ergotropy."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Provides the random-matrix construction of Bures-measure states used in the numerical concentration analysis for the least-informative prior."}],"review_version":1}