{"id":"ad019bf2-2ce7-4067-b20d-6fbfe0ee1453","arxiv_id":"2602.21640","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"For trapped attractive Fermi gases in 1D and 2D, as N grows the ground-state energy approaches the Thomas-Fermi energy, and ground states converge via Husimi functions.","lead":"This paper proves that the ground-state energy of a trapped Fermi gas with attractive short-range interactions converges to a Thomas-Fermi energy in the large-particle limit, in one and two dimensions. The proof handles attractive interactions—where standard repulsive-potential methods fail—and also proves convergence of the many-body states through their Husimi functions.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"A priori bound in Lemma 3.2 mis-scales for d=2; corrected exponent breaks parameter choices in (3.130), leaving Theorem 1.8 (d=2) unsupported.","rationale":"The reader identified Lemma 3.2 as the weakest assumption and noted that every subsequent error estimate depends on it. However, the reader did not pinpoint the concrete scaling error: equation (3.14) miscomputes the L^p norm of w_N for d=2, which changes the exponent in the a priori bound and breaks the parameter choices in (3.130). This is an internal inconsistency, not a disagreement with consensus. It does not affect d=1, but it undermines the central claim for d=2 as written. The issue is potentially fixable by a different parameter choice, so I retain the reader's CONDITIONAL verdict rather than moving to REJECT. The abstract claim about repulsive interactions without a proof remains a separate, secondary concern.","tokens_in":33236,"tokens_out":14857,"duration_ms":112149,"concrete_test":"Recompute the L^{1+d/2} norm of w_N = N^{dβ}w(N^β·) and substitute into (3.13)–(3.14). Then re-derive (3.5) and (3.102) with the corrected exponent. Finally, check whether there exist parameters satisfying the tiling constraints (3.62), (3.65), (3.81) and making all error terms in (3.129) vanish, i.e. τ ≫ N^{4β} (corrected probability term) and τ^2N^{2β}/L^4 ≪ 1 simultaneously for d=2. If no such τ can be chosen, the β range in Theorem 1.8 must be restricted or the proof revised.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The proof of the lower bound depends critically on Lemma 3.2, whose bound (3.5) is used in Lemma 3.21 and in the final parameter choices. In deriving (3.5), equation (3.14) claims ||w_N||_{L^{1+d/2}}^{1+d/2} = N^{βd/2}∫w^{1+d/2}. For w_N = N^{dβ}w(N^β·), the L^p norm scales as ||w_N||_p = N^{dβ(p-1)/p}||w||_p. With p=1+d/2, this gives ||w_N||_p^p = N^{d^2β/2}||w||_p^p, not N^{βd/2}. The identity in (3.14) is only correct for d=1; for d=2 the correct exponent is 2β, not β. Consequently, (3.5) should read O(N^{1+d^2β/2}) = O(N^{1+2β}) for d=2, not O(N^{1+β}). This error propagates to (3.102): P(Ξ^c) ≤ Cτ^{-1}N^{2β} instead of Cτ^{-1}N^{β}. The term N^{dβ}P(Ξ^c) in (3.129) then becomes C N^{4β}τ^{-1} instead of C N^{3β}τ^{-1}. The stated parameter choice (3.130) requires τ ≪ N^{2βd}=N^{4β} (from the τ^2N^{dβ}/L^4 term) but also needs the probability term to vanish; with the corrected exponent, that requires τ ≫ N^{4β}, so the window closes and the error term is O(1), not o(1). The lower bound would only give liminf ≥ ETF - C, not ≥ ETF. A different parameter choice might rescue the proof, but none is provided, so Theorem 1.8 as stated for d=2 is not proven.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies N spin-polarized fermions in d=1,2 with Hamiltonian (1.12), semiclassical scaling ℏ=N^{-1/d}, and short-range attractive interactions w_N=N^{dβ}w(N^β·). It claims E(N)=N E_TF+o(N) (Theorem 1.8) for β<2/[d(2d+1)], and convergence of Husimi functions of approximate ground states to minimizers of the Vlasov energy (Theorem 1.17). The strategy is an upper bound via the Hartree/Lieb variational principle, and a lower bound via Husimi functions, the Diaconis–Freedman theorem, averaging of empirical measures, and an approximate Pauli principle. The d=1 case is treated with a relaxed functional due to non-uniqueness/discontinuity of minimizers.","tokens_in":33696,"tokens_out":12667,"duration_ms":92200,"significance":"If correct, the paper would provide a rigorous derivation of the Thomas–Fermi energy for attractive Fermi gases with nonlocal short-range interactions, extending the well-developed repulsive mean-field theory to a physically relevant attractive setting, and including convergence of states. The paper is carefully structured, with many detailed estimates and a self-contained treatment of the 1D minimizer problem. However, the d=2 proof contains a scaling error that propagates through the a priori bounds and parameter choices; as it stands, Theorem 1.8 for d=2 is not established.","major_comments":[{"comment":"The scaling in (3.14) is incorrect for d=2. For w_N=N^{dβ}w(N^β·) and p=1+d/2, ||w_N||_p = N^{dβ(1-1/p)}||w||_p, so ||w_N||_p^{1+d/2} = N^{βd^2/2}||w||_p^{1+d/2}, not N^{βd/2}. Thus (3.16) and (3.5) should be O(N^{1+βd^2/2}), which is O(N^{1+2β}) for d=2. This changes (3.20)-(3.21) and propagates to Lemma 3.3: for d=2 the interaction error in (3.39)-(3.42) becomes O(N^{1+3β}√ℏ_x), requiring ℏ_x≪N^{-6β}. Compatibility with ℏ_xℏ_p=ℏ^2=N^{-1} and ℏ_p≪1 only allows ℏ_x≫N^{-1}, so the lemma fails for β≥1/6. This is a load-bearing error, not a typo.","section":"§3.1, Eq. (3.14), Lemma 3.2"},{"comment":"With the corrected Lemma 3.2, the Markov bound (3.102) becomes P(Ξ^c) ≤ C τ^{-1} N^{βd^2/2}, so the last error term in (3.129) is N^{dβ}P(Ξ^c) ≤ C N^{βd(1+d/2)} τ^{-1}, i.e. N^{4β}τ^{-1} for d=2. The parameter choice (3.130) then requires τ≫N^{4β} (to make this error vanish) and τ≪N^{4β} (from the τ^2N^{dβ}/L^4 term), so the window closes. The lower bound gives at best liminf ≥ E_TF - C, not ≥ E_TF. Hence Proposition 3.1 and Theorem 1.8 for d=2 are not proved as stated, and Theorem 1.17 inherits the gap through (4.1).","section":"§3.5, Eqs. (3.102), (3.129), (3.130)"}],"minor_comments":[{"comment":"The abstract states that the results 'extend to the case of a repulsive interaction of positive Fourier transform', but no theorem, proposition, or example in the text addresses repulsive interactions. The proof relies on the attractive sign and on Assumption 1.7 (Iw < cTF in d=2). This claim should be removed or substantiated.","section":"Abstract"},{"comment":"The prefactor in the definition of the semiclassical Fourier transform appears to read (2πℏx)^{-d/2}; it should presumably be (2πℏ)^{-d/2}.","section":"Definition 1.11, Eq. (1.37)"},{"comment":"Typo: 'Additionnal material' should be 'Additional material'.","section":"Appendix title"},{"comment":"The 1-Wasserstein distance is defined with sup over ∥φ∥_{Lip}≤1; the Lipschitz seminorm is usually denoted |φ|_{Lip}. Please clarify notation.","section":"Definition 4.3"}],"recommendation":"major_revision","confidential_remarks":"The paper is promising and the d=1 part appears solid, but the d=2 claim is currently unsupported due to the scaling error in Lemma 3.2 and its consequences. I would not reject outright: it may be repairable with a sharper a priori estimate or a restricted range of β, but the revision must address this before the main theorem can be accepted."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Key thing to know: the d=2 half of Theorem 1.8 is not supported as written. In Lemma 3.2, the scaling identity in (3.14) is wrong for d=2. For w_N = N^{dβ} w(N^β ·), the L^p norm satisfies ||w_N||_p^p = N^{dβ(p−1)} ||w||_p^p. With p=1+d/2, this is N^{d^2β/2}, not N^{dβ/2}. So the a priori bound (3.5) should be O(N^{1+d^2β/2}), not O(N^{1+βd/2}). That exponent flows into (3.102) and then into (3.129), where the probability term becomes N^{4β}/τ for d=2 instead of N^{3β}/τ. With the parameter choices in (3.130), the window τ ≫ N^{4β}, τ ≪ N^{4β} closes. So the d=2 lower bound, and hence Theorem 1.8 for d=2, is not proven. A different choice of scales might rescue it, but none is given. The d=1 case survives the correction, since d=1 makes the two exponents agree.\n\nWhat is genuinely good: the paper treats attractive short-range interactions, a regime that the usual repulsive methods (Onsager, dropping positive terms) don't touch. The upper bound via Lieb's variational principle is clean. The lower-bound strategy — Diaconis–Freedman, then averaging empirical measures to restore an approximate Pauli principle — is substantial and mostly self-contained. There is no circularity and nothing is fitted; the Thomas–Fermi functional is defined from the model constants. The d=1 minimizer analysis in the appendix is careful.\n\nTwo other issues, in proportion. The abstract's last sentence claims extension to repulsive interactions with positive Fourier transform; no theorem or proof for that appears anywhere in the text. Add the result or delete the claim. Minor: the approximate Pauli principle is imported from [31]; that is fine as a tool, but the d=1 assumption on level sets of V (Assumption 1.6) is easy to miss and does real work.\n\nWho this is for: specialists in mathematical quantum many-body theory. It deserves a serious referee — the strategy is real and the d=2 bug looks fixable — but the paper as posted overreaches on both the d=2 statement and the repulsive extension. My vote: send to peer review, with instructions to focus on Lemma 3.2 and the parameter windows in Section 3.5, and to require the repulsive claim to be either proven or removed.","headline":"The d=1 result looks plausible, but Lemma 3.2 mis-scales in d=2 and the abstract promises a repulsive case that isn't proven; the paper needs a fix before it can be used as stated.","tokens_in":34152,"tokens_out":7633,"would_cite":false,"duration_ms":59764,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["81Q10","81V70","46N50","82B10"],"pacs":["05.30.Fk","03.65.Sq","71.10.Ca"],"model":"deepseek-v4-flash","headline":"Ground-state energy of a trapped attractive Fermi gas converges to the Thomas-Fermi energy in one and two dimensions, and approximate ground states converge to Vlasov minimizers.","keywords":["Fermi gas","attractive interaction","semi-classical limit","Thomas-Fermi functional","Husimi functions","Diaconis-Freedman theorem","ground-state energy","Vlasov energy"],"falsifier":"If one could construct a sequence of N-fermion states with energy of order N but with one-body energy growing faster than N^{1+βd/2}, the a priori bound would be false, and the lower bound argument would fail. Numerically, one could simulate the N-body Schrödinger equation for d=1 with a short-range attractive potential and check whether the ground-state energy divided by N approaches the Thomas-Fermi value for β up to 1/d.","tokens_in":33122,"feed_emoji":"⚛️","tokens_out":1217,"duration_ms":13476,"temperature":0.7,"pith_summary":"This paper studies a trapped gas of N spin-polarized fermions with attractive short-range interactions in one or two dimensions. It shows that, for a certain scaling of the interaction range, the ground-state energy is asymptotically equal to N times the Thomas-Fermi energy, a simple functional of the spatial density. It also proves that the Husimi functions (quantum phase-space distributions) of approximate ground states converge to minimizers of the corresponding Vlasov energy. The result extends the known mean-field semi-classical limit to attractive interactions, where standard methods for repulsive potentials fail. A key step is a quantitative mean-field approximation using the Diaconis-Freedman theorem, after averaging empirical measures to restore the Pauli principle.","feed_headline":"Attractive Fermi gas energy converges to Thomas-Fermi","feed_subtitle":"New proof covers 1D and 2D trapping with short-range attraction, and shows ground states become Vlasov minimizers.","key_machinery":"The argument rests on the Thomas-Fermi functional (1.20) as the effective energy, and on the Husimi functions m^{(k)}_{Ψ_N} (smoothed phase-space densities) for the lower bound. The lower bound rewrites the N-body energy as an integral over empirical measures using the Diaconis-Freedman theorem, then averages these measures over small phase-space cells to restore the Pauli principle, and finally compares the resulting energy to the Thomas-Fermi functional. The parameter β controlling the interaction range is constrained by the validity of the a priori one-body energy bound in Lemma 3.2.","core_discovery":"The central claim is that for d=1,2 and 0<β<2/(d(2d+1)), the ground-state energy E(N) of the Hamiltonian with N fermions, a confining potential V, and attractive interaction w_N = N^{dβ} w(N^β·) satisfies E(N)=N E_TF + o(N), where E_TF is the minimum of the Thomas-Fermi functional c_TF ∫ρ^{1+2/d} + ∫Vρ - I_w∫ρ². Furthermore, the one-body Husimi functions of approximate ground states converge, up to extraction, to a probability measure on phase-space measures that is concentrated on minimizers of the Vlasov energy; the limiting measures satisfy the Pauli principle 0≤m≤(2π)^{-d}. The proof uses an upper bound via Lieb's variational principle and a lower bound built from semi-classical approxim","pith_inferences":["The restrictive condition β<2/[d(2d+1)] appears technical, and one might expect the optimal threshold to be β<1/d; the author's own remark suggests improvement is plausible with stronger a priori bounds.","The 1D case has non-unique Thomas-Fermi minimizers (e.g., double-well potentials can populate one well), so any quantitative rate or uniqueness claim would require additional structure.","The method could likely be adapted to spin-1/2 fermions, with the spin degeneracy merely altering the constant c_TF, as the author notes.","A testable extension is to compute the next-order correction to the energy (order N^{1-?}) to see whether the Thomas-Fermi functional is indeed the full leading term in this attractive scaling."],"forward_implications":["The ground-state energy of a trapped attractive Fermi gas is asymptotically given by a simple density functional, enabling quantitative predictions for experiments with Feshbach-tuned attractive interactions.","The convergence of Husimi functions means that quantum ground states become semiclassical measures supported on the phase-space minimizer, validating the Vlasov equation description for such systems.","The repulsive case with positive Fourier transform is also covered, showing the method is not specific to attraction.","The result extends the semi-classical limit to interaction scalings beyond the mean-field regime, where the interaction becomes local in the limit."],"fun_headline_variants":["1D/2D attractive Fermi gas energy hits Thomas-Fermi at large N","Fermi gas ground-state energy converges to Thomas-Fermi in 1D/2D","Attractive fermions: Husimi functions show Vlasov limit","Semi-classical limit proved for attractive Fermi gas in low dimensions"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The a priori bound (3.5) that energy-bounded states have one-body energy at most N^{1+βd/2} is the load-bearing estimate; if it fails, the lower bound construction collapses and the allowed range of β shrinks toward zero.","fun_headline_variants_meta":{"raw":{"variants":["1D/2D attractive Fermi gas energy hits Thomas-Fermi at large N","Fermi gas ground-state energy converges to Thomas-Fermi in 1D/2D","Attractive fermions: Husimi functions show Vlasov limit","Semi-classical limit proved for attractive Fermi gas in low dimensions"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000538,"raw_usage":{"total_tokens":2380,"prompt_tokens":669,"completion_tokens":1711,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":413,"completion_tokens_details":{"reasoning_tokens":1625}},"tokens_in":413,"tokens_out":1711,"duration_ms":10894,"temperature":1.0,"reasoning_tokens":1625,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-02T20:57:11.388837+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"If one could construct a sequence of N-fermion states with energy of order N but with one-body energy growing faster than N^{1+βd/2}, the a priori bound would be false, and the lower bound argument would fail. Numerically, one could simulate the N-body Schrödinger equation for d=1 with a short-range attractive potential and check whether the ground-state energy divided by N approaches the Thomas-Fermi value for β up to 1/d.","supporting_citations":[],"review_version":1}