{"id":"5a5d95a0-4fd7-42d9-a9c2-631584bc3acb","arxiv_id":"2501.15803","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"For trapped Bose-Einstein condensates in the Gross-Pitaevskii regime, the exponential moment <ψ_N, exp(κ N_⊥)ψ_N> is uniformly bounded in N for small κ>0, giving exponential decay of the excitation probability.","lead":"This paper proves that in a trapped Bose gas, the number of particles outside the condensate is exponentially suppressed on low-energy states, extending a result previously known only for gases on a torus. The proof combines earlier renormalization methods with new commutator estimates for the exponential of the excitation number operator.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Deferred proof of Lemma 4.2 and Propositions 4.4–4.6 is load-bearing; position-space extension of [27] is asserted, not demonstrated, so Theorem 1.2 is not fully established from this manuscript.","rationale":"The reader's conditionality is justified: the main technical ingredient that is genuinely new relative to [11] is Lemma 4.2, and the paper explicitly defers its proof. The rest of the argument from Proposition 3.2 to Theorem 1.2 is clean and checkable. The concern is not that the result is false but that the manuscript does not contain enough detail to verify the load-bearing commutator estimates in an inhomogeneous, non-compact setting. This is a normal reason to require revisions before full acceptance. My independent reading agrees with the reader's weakest_assumption, and the proposed concrete test directly targets the unproved step.","tokens_in":22087,"tokens_out":2919,"duration_ms":28281,"concrete_test":"Write out a full position-space proof of Lemma 4.2 for η in (3.4), using Lemma 3.1 and the exponential decay of φ_GP. In particular, derive (4.9) by expanding [e^{λN}, d_{η,x}] through the series (4.1)–(4.5) and verifying that the double commutator produces the factor λ^2 with all R^3 integrals finite, matching the stated right-hand side. If the derivation requires extra assumptions on V_ext or η, state them explicitly. As a secondary check, provide a complete proof of at least one of Propositions 4.4–4.6, confirming the error bounds (4.42) or (4.52). If both checks pass without new hypotheses, the concern is resolved.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The central claim rests on the commutator bound (3.8) of Proposition 3.2. Its proof depends on Lemma 4.2, which transfers [27, Lemma 2.4] from the torus to R^3 for the kernel η in (3.4). This kernel is not translation-invariant: it contains Q⊗Q acting on φ_GP⊗φ_GP times a radial function, with φ_GP a spatially decaying minimizer. On the torus, Fourier modes decouple, and the discrete momentum bounds simplify the λ^2 double-commutator estimates. In R^3 the spatial variation of φ_GP and the unbounded domain introduce terms (derivatives hitting φ_GP, boundary contributions at infinity) that are not addressed. The paper states 'up to minor technical modifications' and skips the details. Propositions 4.4–4.6 are likewise asserted by reference to [11,12,27] without proof; they control the error terms E_N^(1)–E_N^(4) under the exponential weight e^{κN/2}. If any of these deferred estimates fail, the bound (3.8) and hence Theorem 1.2 do not follow. No in-text evidence indicates the bounds are false, but they are not established in this manuscript.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper treats N trapped bosons in R^3 interacting via N^2 V(N(x-y)) and proves an exponential bound for the number of excitations out of the Gross-Pitaevskii minimizer: for low-energy spectral subspaces, ⟨ψ_N, exp(κ N_{⊥φ_GP}) ψ_N⟩ ≤ C uniformly in N (Theorem 1.2). The proof proceeds in three stages. Section 2 gives a complete proof of the analogous exponential bound for translation-invariant mean-field bosons on the torus. Section 3 reduces the Gross-Pitaevskii problem to Proposition 3.2: after conjugating the excitation Hamiltonian by a generalized Bogoliubov transformation generated by the kernel η in (3.4), the renormalized Hamiltonian G_N satisfies an energy lower bound (3.7) and an exponential commutator bound (3.8). The main theorem follows from these bounds by a Grönwall argument. Section 4 analyzes G_N term by term, but several decisive estimates (Lemma 4.2 and Propositions 4.4–4.6) are asserted with references to [11, 12, 27] and 'minor technical modifications' rather than proved.","tokens_in":22334,"tokens_out":9716,"duration_ms":86412,"significance":"If the proof is completed, Theorem 1.2 is a natural and valuable extension of [27] from translation-invariant settings to trapped gases, and it strengthens the O(1) moment bounds of [11, 12] to exponential moment control. The paper is cleanly organized, gives a self-contained mean-field model, identifies the external inputs (the lower bound (1.8), the energy asymptotics, and the kernel bounds of Lemma 3.1), and does not appear to assume the conclusion. The main obstacle is that the technical core supporting Proposition 3.2 is deferred; as it stands the manuscript does not fully establish its central claim.","major_comments":[{"comment":"Lemma 4.2 is the principal technical input for the commutator bound (3.8), but its proof is not given: the text states that [27, Lemma 2.4] applies in position space on R^3 'up to minor technical modifications' and then 'we skip the details.' This is load-bearing because the kernel η in (3.4) is not translation-invariant: it contains Q⊗Q and the spatially varying, exponentially decaying φ_GP, so the discrete-momentum arguments of [27] do not directly transfer; one must control commutators in which derivatives hit φ_GP and the kernel η, as well as boundary terms at infinity. I ask for a complete proof, or a precise theorem statement in a published reference with hypotheses that match the present setting, before (3.8) can be used.","section":"4.1, Lemma 4.2"},{"comment":"The estimates for G_N^(1) through G_N^(4) are asserted rather than proved: Proposition 4.4 is justified by 'we omit further details and refer to [11, Section 4.2] ... [27, Section 3.2]', Proposition 4.5 by 'we omit the details here', and Proposition 4.6 by 'We again skip the details'. These propositions control both the expectation values and the exponential commutator bounds for the error terms E_N^(1)–E_N^(4); they feed directly into (4.56)–(4.57) and hence into the lower bound (3.7) and the exponential bound (3.8). Since the trapping potential, the kernel η, and the generalized Bogoliubov transformation differ from the torus setting, the asserted 'same arguments' need to be written out or cited to a theorem that covers this exact configuration.","section":"4.3–4.5, Propositions 4.4, 4.5, 4.6"},{"comment":"In the proof of Proposition 3.2, the scattering term displayed in (4.58) is claimed to be bounded by C(N_{⊥φ_GP}+1) with no derivation. This bound is needed for the energy lower bound (3.7), which is itself used in the proof of Theorem 1.2. Please provide the argument (including the treatment of the boundary terms from the cutoff at |x−y|=ℓ), or give a precise reference to a statement with the same trapping setup.","section":"4.6, around Eq. (4.58)"}],"minor_comments":[{"comment":"There are typos in the abstract ('Gross-Pit aevskii', 'Bose-Einst ein') and in Remark 1.5 ('everty 0 ≤ n ≤ N' should be 'every 0 ≤ n ≤ N').","section":"Abstract, Remark 1.5"},{"comment":"The parentheses in `(N + 1)n+6)/2` are unbalanced, and in (4.8) the term `‖η‖2‖ay‖(N + 1)(n+4)/2ξ‖` appears to be missing a norm around `ay(N+1)^{(n+4)/2}ξ`; please check all occurrences.","section":"Eqs. (4.8), (4.12), (4.13)"},{"comment":"The function in (4.39) is written as `ω(N·)`, but the function defined earlier is `wℓ = 1 − fℓ`; Section 4.5 uses `w(N·)`. Please unify the notation.","section":"Eq. (4.39) and Section 4.5"},{"comment":"In the derivation of (2.3), the vector `(H_N^mf − E_N^mf)ψ_N / ||(H_N^mf − E_N^mf)ψ_N||` is placed in Q_ζ. If the denominator vanishes, this step should be justified by a limiting argument or by treating ground states separately.","section":"Proof of Theorem 2.1"},{"comment":"The sentence 'The proof of Lemma 4.2 goes bac to NR' is an incomplete attribution; please give the full reference or remove it.","section":"Section 4.1"},{"comment":"The parentheses in the displayed formula for G_N^(3) are mismatched; it should read `−√N [b(coshη(h_N)) + b∗(sinhη(h_N)) + h.c.]`.","section":"Eq. (4.49)"}],"recommendation":"major_revision","confidential_remarks":"To the editor: the main mathematical input for the trapped setting is drawn from a chain of recent papers by overlapping authors, including the preprint [27] and the master’s thesis [4]. The manuscript is transparent about this, but the amount of deferred material is too large for a standalone proof. If the authors can supply the missing proofs in a revision, I would be happy to reconsider; otherwise the paper remains more of a research announcement than a complete proof."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The paper claims exponential control of the number of excitations for trapped BECs in the Gross-Pitaevskii regime, generalizing [27] from the torus to R^3 with an external potential. If the main theorem holds, it is a natural and useful extension: it gives the exponential moment bound that feeds into large-deviation and moment-generating-function results. The statement itself is new, and the high-level strategy is coherent. The mean-field analogue in Section 2 is proved completely and is a clean illustration of the commutator method. The reduction of Theorem 1.2 to Proposition 3.2 is also clearly laid out. The paper is honest about what it does not prove: Remark 1.3 mentions the master thesis [4], and the text repeatedly states 'up to minor technical modifications' and 'we omit the details'.\n\nThe soft spot is exactly this: the load-bearing technical core is deferred. Lemma 4.2 transfers the commutator error bounds of [27, Lemma 2.4] from the torus (discrete momentum space) to position space R^3 for the kernel eta in (3.4). That kernel is not translation-invariant; it contains Q x Q acting on phi_GP x phi_GP times a radial function, with phi_GP decaying but spatially varying. On the torus, Fourier modes decouple and the discrete momentum bounds simplify the double-commutator estimates. In R^3, derivatives hitting phi_GP and the lack of compactness could introduce extra terms. The paper asserts these are 'minor technical modifications' and gives no derivation. Propositions 4.4, 4.5, and 4.6, which control the errors E_N^(1) through E_N^(4) under the exponential weight, are likewise asserted by reference to [11,12,27] without proof. These are not peripheral estimates; the commutator bound (3.8) of Proposition 3.2 depends on them directly, and without (3.8), Theorem 1.2 does not follow.\n\nTo be clear: I see no evidence that the bounds are false, and the paper's own structure makes the dependency explicit. The author overlap with [11,12,27] is not a problem in itself; the new statement is clearly theirs. But a rigorous proof needs either a supplement spelling out Lemma 4.2 and the propositions, or precise pointers to where each statement appears verbatim in the cited works. As it stands, the paper is a credible research announcement with a proof sketch, not a complete proof.\n\nWho is this for? Researchers working on rigorous BEC theory and large deviations for Bose gases. They will want to read it, but they should treat the main theorem as conditional pending the deferred estimates. I would send it to a serious referee, with instructions to check whether the claimed 'minor modifications' are indeed minor, especially for the R^3 kernel. If the referee can fill the gaps, the result is a solid contribution. If not, the paper needs substantial revision.\n\nRecommendation: engage with the paper, but require the technical details before accepting the theorem as established.","headline":"Trapped BEC exponential control is a credible and clearly explained generalization of the torus result, but the load-bearing commutator estimates are deferred in the text, so the main theorem is not fully established from this manuscript.","tokens_in":22911,"tokens_out":2835,"would_cite":true,"duration_ms":25564,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["35Q55","81Q10","82B10"],"pacs":[],"model":"deepseek-v4-flash","headline":"A trapped dilute Bose gas has exponentially small probability of large condensate depletion in low-energy states.","keywords":["Bose-Einstein condensation","Gross-Pitaevskii regime","excitation number operator","exponential moment bound","generalized Bogoliubov transformation","large deviations","trapped Bose gas","low-energy spectral subspace"],"falsifier":"Write out the proof of Lemma 4.2 for the kernel $\\eta$ of (3.4) and check the claimed bounds on $[e^{\\lambda\\mathcal{N}}, d_{\\eta,x}]$ and on the double commutators; if any of these bounds fails for a potential satisfying Assumption 1.1, the commutator estimate (3.8) and with it the proof of Theorem 1.2 breaks. A less structural check is to compute, for a sequence of $N$, the exponential moment $\\langle \\psi_N, e^{\\kappa N_{\\perp\\phi_{\\mathrm{GP}}}}\\psi_N\\rangle$ for low-energy states of a trapped gas and look for divergence.","tokens_in":21776,"feed_emoji":"⚛️","tokens_out":12727,"duration_ms":109606,"temperature":0.7,"pith_summary":"This paper proves exponential control of the number of excited particles in a trapped three-dimensional Bose gas in the Gross-Pitaevskii regime: for any low-energy state, the expectation value of $\\exp(\\kappa N_{\\perp\\phi_{\\mathrm{GP}}})$ is bounded uniformly in the particle number $N$, provided $\\kappa$ is small enough. By Markov's inequality this means the probability of finding at least $n$ particles outside the condensate decays like $e^{-\\kappa n}$, which is much stronger than the polynomial decay that follows from earlier moment bounds. The result extends to trapped gases a theorem that was known for translation-invariant gases on a torus, bringing the exponential bound closer to experimental setups with an external trap. The proof works directly on spectral subspaces of the full Hamiltonian, not only on eigenstates, and its core is a commutator estimate for a renormalized excitation Hamiltonian.","feed_headline":"Low-energy trapped Bose gas: excitations decay exponentially","feed_subtitle":"The chance of n particles leaving the condensate is bounded by e^{-κn}, uniformly in N.","key_machinery":"The mechanism is a renormalized excitation Hamiltonian built from a generalized Bogoliubov transformation. The transformation $e^{B(\\eta)}$ is generated by a two-particle correlation kernel $\\eta$ that encodes the short-scale scattering correlations between particles and is constructed from the scattering correction $w_\\ell$, a high-momentum cut-off, and projections orthogonal to the condensate. Conjugating the excitation Hamiltonian $L_N$ with $e^{B(\\eta)}$ gives $G_N$, in which dangerous terms cancel and the relevant object is the excitation number operator $\\mathcal{N}$, unitarily equivalent to $N_{\\perp\\phi_{\\mathrm{GP}}}$. The load-bearing estimate is the commutator bound (3.8): for small $\\kappa$, $|\\langle\\xi, e^{\\kappa\\mathcal{N}/2}[G_N,e^{\\kappa\\mathcal{N}/2}]\\xi\\rangle| \\le C\\kappa \\langle\\xi,(\\mathcal{H}_N+\\mathcal{N})e^{\\kappa\\mathcal{N}}\\xi\\rangle$, with $\\mathcal{H}_N$ a renormalized kinetic-plus-interaction operator. This bound, combined with the lower bound $G_N \\ge N E_{\\mathrm{GP}}(\\phi_{\\mathrm{GP}})+c(\\mathcal{H}_N+\\mathcal{N})-C$ and a bootstrap comparing $\\langle e^{\\kappa\\mathcal{N}}\\rangle$ to its derivative, yields the uniform exponential moment bound. The same commutator bootstrap is first illustrated in the simpler mean-field translation-invariant setting, where the computation is explicit.","core_discovery":"The central claim is Theorem 1.2: under mild assumptions on the nonnegative interaction $V$ and the trapping potential $V_{\\mathrm{ext}}$, for any fixed energy window $\\zeta$ above the ground state energy $E_N$, every normalized $N$-particle wave function $\\psi_N$ with energy at most $E_N+\\zeta$ satisfies $\\langle\\psi_N,\\exp(\\kappa N_{\\perp\\phi_{\\mathrm{GP}}})\\psi_N\\rangle \\le C$ for all sufficiently small $\\kappa>0$, with $C$ independent of $N$ and $\\psi_N$. Here $\\phi_{\\mathrm{GP}}$ is the unique positive minimizer of the Gross-Pitaevskii functional and $N_{\\perp\\phi_{\\mathrm{GP}}}$ counts particles orthogonal to the condensate wave function. The analogue on the unit torus was proved in [27]; this paper adapts the argument to the position-space setting of a trap, where momentum space is not discrete. The exponential moment bound yields the tail estimate $\\mathbb{P}[N_{\\perp\\phi_{\\mathrm{GP}}} \\ge n] \\le e^{-\\kappa n}$ for low-energy states, so the depletion of the condensate has subexponential probability of being large. The paper also provides an illustrative mean-field version of the argument and states that it expects a full moment-generating-function and large-deviation refinement in the trapped setting.","pith_inferences":["The proof's only deferred step is the position-space transfer of [27, Lemma 2.4] into Lemma 4.2; a reader extending this work should verify that transfer first, since a failure there would not disprove exponential control but would invalidate this proof.","A natural next step, which the paper leaves open, is to derive an explicit limiting moment generating function for $N_{\\perp\\phi_{\\mathrm{GP}}}$ in the trap, following the torus route of [31] using [12,28].","Because the commutator estimate is formulated on Fock space and does not use stationarity, the same strategy may yield exponential control of the depletion along the Gross-Pitaevskii time evolution, not just in equilibrium.","The constant $\\kappa$ in Theorem 1.2 is not made quantitative; computing its dependence on the trap and the energy window $\\zeta$ would be a concrete numerical test of the bound."],"forward_implications":["For any low-energy state of a trapped Gross-Pitaevskii gas, the probability of finding at least $n$ particles outside the condensate decays at least as $e^{-\\kappa n}$, uniformly in the particle number $N$.","Expanding the exponential shows the uniform moment bound $\\langle \\psi_N, N_{\\perp\\phi_{\\mathrm{GP}}}^k \\psi_N\\rangle \\le C_k$ for every fixed $k$, recovering and strengthening the optimal-rate condensation estimates of [11,12,25].","The bound holds on the whole spectral subspace of energies up to $E_N+\\zeta$, so it applies to approximate ground states and low-energy ensembles, not only to individual eigenfunctions.","The mean-field analogue proved in Section 2 shows the same commutator bootstrap works without the correlation-kernel renormalization, isolating the role of the renormalization in the Gross-Pitaevskii regime."],"supporting_citations":[{"why":"Supplies the essential lower bound (1.8) on the trapped Hamiltonian and the renormalization framework from which $G_N$ is built.","marker":"[11]"},{"why":"Establishes the uniform moment bounds for trapped bosons and the analysis of the excitation Hamiltonian that this proof refines into exponential control.","marker":"[12]"},{"why":"Provides the exponential commutator method and Lemma 2.4, whose error estimates Lemma 4.2 transfers from the torus to position space.","marker":"[27]"},{"why":"Gives the moment-generating-function analysis for the translation-invariant setting, including the lemma adapted to prove Lemma 3.3.","marker":"[31]"},{"why":"Defines the Gross-Pitaevskii functional and the minimizer properties used throughout, including Theorem A.1.","marker":"[24]"},{"why":"Proves the optimal rate of condensation $\\langle N_{\\perp\\phi_{\\mathrm{GP}}}\\rangle=O(1)$ used together with [11,28,12] to identify the ground state energy $E_N=N E_{\\mathrm{GP}}(\\phi_{\\mathrm{GP}})+O(1)$.","marker":"[25]"},{"why":"Establishes the nested-commutator expansion and boundedness properties of the generalized Bogoliubov transform that Section 4 uses.","marker":"[10]"}],"fun_headline_variants":["Exponential control of excitations in trapped BEC","Trapped BEC: exponential decay of excitations","Exponential suppression of condensate depletion in trapped BEC","Trapped Bose gas: excitations drop off exponentially","Exponential bound on excitations for trapped BEC"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the commutator error estimates for the renormalized Hamiltonian, which the proof takes over from the translation-invariant torus setting 'up to minor technical modifications,' remain valid in position space on $\\mathbb{R}^3$; the paper asserts this transfer in Lemma 4.2 and does not derive it.","fun_headline_variants_meta":{"raw":{"variants":["Exponential control of excitations in trapped BEC","Trapped BEC: exponential decay of excitations","Exponential suppression of condensate depletion in trapped BEC","Trapped Bose gas: excitations drop off exponentially","Exponential bound on excitations for trapped BEC"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000536,"raw_usage":{"total_tokens":2539,"prompt_tokens":876,"completion_tokens":1663,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":492,"completion_tokens_details":{"reasoning_tokens":1586}},"tokens_in":492,"tokens_out":1663,"duration_ms":11523,"temperature":1.0,"reasoning_tokens":1586,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-10T13:56:31.583376+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Write out the proof of Lemma 4.2 for the kernel $\\eta$ of (3.4) and check the claimed bounds on $[e^{\\lambda\\mathcal{N}}, d_{\\eta,x}]$ and on the double commutators; if any of these bounds fails for a potential satisfying Assumption 1.1, the commutator estimate (3.8) and with it the proof of Theorem 1.2 breaks. A less structural check is to compute, for a sequence of $N$, the exponential moment $\\langle \\psi_N, e^{\\kappa N_{\\perp\\phi_{\\mathrm{GP}}}}\\psi_N\\rangle$ for low-energy states of a trapped gas and look for divergence.","supporting_citations":[{"cited_title":"Brennecke, B","cited_arxiv_id":null,"evidence_quote":"Supplies the essential lower bound (1.8) on the trapped Hamiltonian and the renormalization framework from which $G_N$ is built."},{"cited_title":"Brennecke, B","cited_arxiv_id":null,"evidence_quote":"Establishes the uniform moment bounds for trapped bosons and the analysis of the excitation Hamiltonian that this proof refines into exponential control."},{"cited_title":"Exponential bounds of the condensation for dilute Bose gases","cited_arxiv_id":"2307.10622","evidence_quote":"Provides the exponential commutator method and Lemma 2.4, whose error estimates Lemma 4.2 transfers from the torus to position space."},{"cited_title":"Generating function for quantum depletion of Bose-Einstein condensates","cited_arxiv_id":"2410.13576","evidence_quote":"Gives the moment-generating-function analysis for the translation-invariant setting, including the lemma adapted to prove Lemma 3.3."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Defines the Gross-Pitaevskii functional and the minimizer properties used throughout, including Theorem A.1."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Proves the optimal rate of condensation $\\langle N_{\\perp\\phi_{\\mathrm{GP}}}\\rangle=O(1)$ used together with [11,28,12] to identify the ground state energy $E_N=N E_{\\mathrm{GP}}(\\phi_{\\mathrm{GP}})+O(1)$."},{"cited_title":"Brennecke, B","cited_arxiv_id":null,"evidence_quote":"Establishes the nested-commutator expansion and boundedness properties of the generalized Bogoliubov transform that Section 4 uses."}],"review_version":1}