{"id":"631a1d8a-d85e-4f06-a3d7-b08c35e55e9e","arxiv_id":"2607.23579","paper_version":1,"verdict":"ACCEPT","confidence":"HIGH","novelty_score":7.0,"correctness_risk":"low","formal_verification":"none","parameter_count":0,"one_line_summary":"Any U(1)-symmetric permutation-invariant fermionic state is completely reconstructed from the particle-number distribution and one uniform-mode occupation per sector, with linearly many observables.","lead":"A fermionic quantum state that is unchanged under swapping modes, and that conserves particle number, is fully fixed by the particle-number histogram plus one collective-mode occupation per sector. The protocol needs only linearly many measurements already available in cold-atom band mapping, so simulators can extract state-level information without full tomography.","discovery_kind":"new_method","skeptic_critique":{"model":"moonshotai/kimi-k3","headline":"No significant objection identified. The load-bearing representation-theoretic step (irreducibility and mutual inequivalence of ∧^q W and ∧^{q-1} W) is a standard result, and the one genuinely delicate case — half filling N=2q where dimensions coincide — is handled by a correct explicit character计算","rationale":"The reader's identified weakest assumption is the right place to look — it is the single point where the argument could silently break (a reducible or equivalent pair of sector representations would leave extra invariant operators and underspecify Π(ρ)). But on inspection that assumption is fully secured inside the paper: irreducibility of ∧^r W is the standard hook-irrep result with an appropriate citation, and the only nontrivial case, N=2q where the dimension-counting argument is unavailable, receives an explicit, correct character computation distinguishing the two irreps (the transposition characters are nonzero and opposite in sign). The remaining steps (Fock-space decomposition ∧^q(triv ⊕ W), Schur's lemma, trace-based coefficient determination) are routine and were verified line by line. The fermionic sign subtlety in defining U_π is absorbed in a phase that cancels in the twirl. The two applications are illustrative rather than load-bearing, and the scope limitations (Π(ρ) vs. ρ; no noise analysis; no shipped code) are acknowledged and do not undercut the formal claim. Since the concern I would have raised is the same one the reader flagged, and it is provably settled in the End Matter, there is no basis for moving the verdict: ACCEPT with HIGH confidence stands. The proposed closure test is cheap (N≤8 exact diagonalization-scale) and would convert the analytical argument into a numerically falsified-or-confirmed statement if anyone doubts the End Matter.","tokens_in":14148,"tokens_out":3091,"duration_ms":98477,"concrete_test":"Direct numerical closure check of Eq. 4 at the delicate point: for N=6 (so half filling q=3, where dim ∧^3 W = dim ∧^2 W = 10) and N=8, generate random U(1)-symmetric density matrices ρ (e.g., Haar-random unitary conjugation of a random number-diagonal state), compute Π(ρ) by explicit twirl over S_N (720/40320 terms), and independently compute the right-hand side of Eq. 4 from measured {p_q, ν_q}. If ‖Π(ρ) − RHS‖_1 exceeds numerical precision (~1e−12), a hidden invariant operator exists and the claim fails; agreement at machine precision across all sectors confirms the Schur's-lemma collapse including the equal-dimension sector.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim (Eq. 4) rests on a short Schur's-lemma argument whose only non-mechanical input is: (i) ∧^r W is irreducible as an S_N-representation for all r, and (ii) ∧^q W and ∧^{q-1} W are inequivalent for 1≤q≤N−1, including the equal-dimension case N=2q. I checked both. (i) The one-particle space decomposes as trivial ⊕ standard, so W is the (N−1)-dimensional standard irrep; its exterior powers ∧^r W are the hook irreps (N−r, 1^r), irreducible for all r — a textbook result the paper correctly cites to Fulton–Harris [41]. (ii) For N≠2q the dimensions differ and inequivalence is trivial. For N=2q, the End Matter computes the character of a transposition τ on ∧^r W as C(N−2,r) − C(N−2,r−1) (Eq. 21), which is correct: τ has eigenvalues +1 (mult. N−2) and −1 (mult. 1) on W, and wedge basis vectors either avoid or contain the −1 eigenvector. At N=2q the two characters are ±[C(2q−2,q) − C(2q−2,q−1)], which is nonzero because C(2q−2,q−1) is the central binomial coefficient, strictly larger than C(2q−2,q). Hence the irreps are inequivalent exactly where the dimension argument fails. The rest of the proof — the CAR automorphism U_π c_i U_π† = c_{π(i)} (well-defined up to an irrelevant phase), the decomposition ∧^q(triv ⊕ W) = ∧^q W ⊕ ∧^{q−1} W (Eq. 8), Schur's lemma killing off-diagonal blocks and scalarizing diagonal blocks, and the coefficient extraction via Tr and Tr(·n_0) (Eq. 12) — is mechanical and correct. The stated scope limits (full tomography only for permutation-invariant ρ; otherwise Π(ρ)) are honestly framed and not overstated. The numerical illustrations (SYK fidelity, Lifshitz kink) support but do not carry the central claim. I find no load-bearing weakness.","agreement_with_reader":"agree"},"referee_report":{"model":"moonshotai/kimi-k3","summary":"The manuscript constructs a fermionic analog of permutationally invariant quantum tomography (PIQT) for states with U(1) particle-number symmetry. The central result, Eq. (4), states that the permutation-symmetrized density matrix Π(ρ) of any U(1)-symmetric N-mode fermionic state is completely determined by the particle-number distribution {p_q} and the conditional uniform-mode occupations {ν_q}, taking the block-scalar form Π(ρ) = Σ_q p_q[(1−ν_q)/C(N−1,q) P_{q,0} + ν_q/C(N−1,q−1) P_{q,1}]. The proof decomposes each q-particle sector into the uniform-mode-empty and uniform-mode-occupied subspaces, which carry the exterior powers ∧^q W and ∧^{q−1} W of the standard (N−1)-dimensional S_N irrep; Schur's lemma kills cross blocks and scalarizes diagonal blocks, and the trace conditions fix the coefficients. For permutation-invariant states this constitutes full tomography from O(N) observables. Two applications are presented: a few-sample construction of the disorder-averaged Gibbs state of the complex SYK model (Fig. 1), and the von Neumann entropy of the symmetrized reduced state across a Lifshitz transition, with the asymptotic formula Eq. (17) derived in the Supplemental Material via full counting statistics and Fisher–Hartwig asymptotics.","tokens_in":14556,"tokens_out":3843,"duration_ms":138972,"significance":"If the result holds — and I believe it does — this is a clean and useful contribution. The central formula (Eq. 4) is a parameter-free, exactly derived identity valid for arbitrary (including interacting, non-Gaussian) U(1)-symmetric fermionic states, not an ansatz or fit. The one genuinely delicate representation-theoretic step — the inequivalence of ∧^q W and ∧^{q−1} W in the equal-dimension case N = 2q — is handled correctly and explicitly in the End Matter via a transposition character computation (Eq. 21), which I verified independently. The required measurements (number-resolved zero-momentum occupation via band mapping) are realistic for current ultracold-atom fermion platforms, which makes the protocol more than a formal curiosity. The Supplemental Material's derivation of Eq. (17) (Toeplitz determinants, Fisher–Hartwig, controlled O(1) bookkeeping) is careful and reproducible in principle. The work extends the PIQT idea of Tóth et al. [33] into a genuinely new setting (itinerant fermions with number conservation), and the SYK example demonstrates a non-obvious use — permutation-orbit amplification of finite disorder samples (Eq. 15) — that is an algebraic identity rather()","major_comments":[{"comment":"The second application rests on Eq. (17), S(Π_A(ρ_A)) = N_A h(ϱ) − (1/2)ln N_A + O(ln ln N_A), whose nonanalyticity at the Lifshitz point is inherited entirely from the filling fraction ϱ = ∫(dk/2π)Θ(−ε_k). But ϱ itself is directly measurable from the momentum distribution — indeed it is a simpler observable than the symmetrized entropy. As written, the manuscript does not establish what state-level information S(Π_A(ρ_A)) contains beyond a nonlinear reparametrization of ϱ. Since this example is one of only two demonstrations and underwrites the abstract's claim that the symmetrized state 'can still encode nontrivial many-body and state-level structure beyond conventional few-body observables,' the authors should clarify what, if anything, the entropy (or other diagnostics of Π(ρ)) reveals that is not already visible in the number statistics {p_q} and the occupations {ν_q} themselves. A","section":"Example 2: Lifshitz transition, Eq. (17)"},{"comment":"The protocol's scalability claim ('number of required observables scales only linearly') concerns observable count, but no analysis is given of the statistical cost: how the reconstruction error in Π(ρ) scales with the number of experimental shots used to estimate p_q and ν_q. The SYK example (Fig. 1) demonstrates robustness to few disorder realizations, but that is sample averaging, not shot noise; ν_q is a conditional average whose estimator variance grows when p_q is small (sectors near q=0 and q=N). Since experimental feasibility is a headline claim of the Letter, at least a brief error-propagation discussion (e.g., fidelity vs. shot number, behavior of rare sectors) is needed for the reader to assess practical cost.","section":"Experimental Implementation / Applications"}],"minor_comments":[{"comment":"The x-axis of Fig. 1(c) is labeled β with tick labels spanning 10^-3 to 10^6, which is inconsistent with panels (a,b) at fixed β=10 and implausible as a physical temperature range; presumably the axis is mislabeled or mis-scaled in the extracted figure. Please check and correct the axis label and range.","section":"Fig. 1(c)"},{"comment":"The fidelity is defined as F(ρ,σ) = Tr√(√ρ σ √ρ), the root (unsquared) convention; many references (including some cited) use the squared convention. Stating the convention explicitly in the caption or text would avoid ambiguity.","section":"Example 1, below Eq. (15)"},{"comment":"In the proof of Eq. (4), the coefficients in Eq. (12) follow from Tr P_{q,0} = C(N−1,q) and Tr P_{q,1} = C(N−1,q−1); one line making these dimensions explicit would make the normalization step fully self-contained (the dimensions appear only later, in the End Matter, Eq. (20)).","section":"Proof of Eq. (4), Eq. (12)"},{"comment":"The CAR automorphism defined by U_π c_i U_π† = c_{π(i)} (Eq. 2) exists and acts on the q-particle sector as ∧^q of the permutation matrix on H_1; a brief remark that no additional fermionic sign twists arise (so that Eq. (8) holds as a representation identity, not just a vector-space decomposition) would preempt a natural reader question.","section":"Preliminaries, Eq. (2)"},{"comment":"When {p_q, ν_q} are estimated from noisy data, the reconstructed operator from Eq. (4) remains positive provided p_q ≥ 0 and 0 ≤ ν_q ≤ 1; a one-sentence remark on enforcing these constraints (or a simple projection) in practice would be helpful.","section":"Experimental Implementation"},{"comment":"Reference [53] to the Supplemental Material cites Refs. [54–58]; the Toeplitz/Fisher–Hartwig inputs are standard, but it would help the reader if the SM stated explicitly which step uses [57] (double-scaling Fisher–Hartwig) versus [56].","section":"Supplemental Material"}],"recommendation":"minor_revision","confidential_remarks":"The core result is mathematically sound; I verified the Schur's-lemma argument including the delicate N=2q inequivalence (the transposition characters are negatives of each other and nonzero, since C(2q−2,q−1) > C(2q−2,q)). My two major comments concern the framing and practical-cost support rather than correctness: the Letter slightly oversells what the symmetrized entropy adds in the Lifshitz example, and the absence of any shot-noise analysis weakens the experimental-feasibility claim. Both are addressable in revision without new results. I noticed the reader's report and stress-test note reached the same conclusion on the central derivation; I agree with them and have nothing to add on that front."},"author_rebuttal":null,"desk_editor":{"model":"grok-4.5","letter":"The one thing worth knowing is that they give an explicit, closed-form reconstruction of the permutation-symmetrized density matrix for any U(1)-symmetric fermionic state: only the particle-number probabilities p_q and the conditional uniform-mode occupations ν_q are needed, and the formula is linear in system size. When the state is already permutation-invariant this is exact tomography; otherwise you get Π(ρ), which still lets you evaluate any permutation-invariant observable.\n\nWhat is new is the fermionic case itself. Qubit PIQT is cited and used as precedent, but the exterior-algebra decomposition (uniform mode plus its orthogonal complement W, then ∧^q W and ∧^{q-1} W) is fermion-specific and not a routine transcription. The proof is short and standard: sectors decouple, permutations preserve n_0, Schur kills the cross blocks once the two representations are irreducible and inequivalent, and the two traces fix the scalars. End Matter correctly handles the only delicate point—half filling, where dimensions coincide—by an explicit transposition character that differs. That step checks out; I do not see a load-bearing hole.\n\nThe two demos are well chosen and do not overclaim. For disorder-averaged complex SYK they show that feeding finite-sample {p_q, ν_q} into the formula effectively folds in the whole permutation orbit, so a few dozen realizations already give high fidelity to the large-sample average. For free-fermion chains they show that the von Neumann entropy of the symmetrized reduced state still carries a non-analytic kink at a Lifshitz transition, with an asymptotic formula that matches numerics. Both illustrate retained diagnostic power without pretending that Π(ρ) is the full state.\n\nSoft spots are minor and scoped honestly. No noise analysis, no error bars on the fidelity curves, no shipped code. The information content of Π(ρ) for generic non-symmetric states is only partially charted; they say so. Experimental accessibility via band mapping is real for cold atoms, but that is the intended audience.\n\nThis is for people who run or analyze fermionic quantum simulators and want state-level diagnostics beyond few-body correlators. The math is reproducible from the text. I would send it to referees and I would bring it to reading group.","headline":"Clean fermionic PIQT formula from O(N) band-mapping data; the Schur argument holds, including at half filling, and the scope is honestly limited to Π(ρ).","tokens_in":15494,"tokens_out":613,"would_cite":true,"duration_ms":10414,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"grok-4.5","headline":"Any permutation-invariant fermionic state with particle-number symmetry is fully fixed by the particle-number histogram and one collective-mode occupation per sector.","keywords":["quantum state tomography","permutation invariance","fermions","U(1) particle-number symmetry","ultracold atoms","Sachdev-Ye-Kitaev model","Lifshitz transition","band mapping"],"falsifier":"Exhibit a U(1)-symmetric fermionic state for which two different permutation-symmetrized density matrices share the same {p_q, ν_q}, or show that the characters of the two exterior-power representations of a transposition agree for some half-filled N=2q, contradicting the End Matter calculation that underpins Schur’s lemma.","tokens_in":15128,"feed_emoji":"⚛️","tokens_out":1022,"duration_ms":26107,"temperature":0.7,"pith_summary":"Full quantum state tomography usually needs exponentially many measurements, and fermionic simulators cannot freely rotate every local basis. This paper shows that if the state is invariant under permuting the modes and conserves total particle number, those barriers drop away: the entire density matrix is fixed by the probabilities of each total particle number and, inside each number sector, the occupation of a single uniform collective mode. Both quantities are already readable from band-mapping shots in ultracold-atom experiments, and their number grows only linearly with system size. Even when the true state is not permutation-invariant, the same data reconstruct the symmetrized component, which still carries many-body structure beyond ordinary few-body correlators. The authors demonstrate the reconstruction on interacting SYK states and on free-fermion chains across a Lifshitz transition, arguing that information-theoretic diagnostics become realistic for itinerant fermionic matter.","feed_headline":"Fermion tomography needs only counts and one mode","feed_subtitle":"Permutation-symmetric fermionic states are fixed by particle-number statistics plus a single collective occupation.","key_machinery":"The reconstruction identity (Eq. 4): after decomposing each q-particle sector into uniform-mode-empty and uniform-mode-occupied subspaces, permutation symmetrization collapses every block by Schur’s lemma to a scalar multiple of the corresponding projector, leaving only the two numbers p_q and ν_q per sector.","core_discovery":"For every fermionic density matrix that respects U(1) particle-number symmetry, the permutation-symmetrized state is completely determined by the sector probabilities p_q and the conditional uniform-mode occupations ν_q. Explicitly, Π(ρ) is a classical mixture, over particle number q, of the two projectors onto the subspaces in which the uniform mode is empty or occupied, with weights fixed solely by p_q and ν_q. When the original state is already permutation-invariant, this is exact tomography from linearly many observables.","pith_inferences":["If the representation-theoretic collapse is the only obstruction, analogous linear protocols should exist for other mode groups (e.g., lattice translations or point-group symmetries) once the corresponding irreps are classified.","The fact that a Lifshitz nonanalyticity survives full symmetrization suggests that topology-sensitive entanglement features may often live in the totally symmetric sector and therefore be experimentally cheaper than expected.","Combining the reconstruction with randomized global unitaries that preserve particle number could enlarge the set of extractable diagnostics while still avoiding site-resolved control."],"forward_implications":["Permutation-invariant fermionic states can be tomographed with O(N) band-mapping observables already available in optical-lattice experiments.","Expectation values of every permutation-invariant observable equal those of the reconstructed Π(ρ), even when the true state is not symmetric.","Disorder-averaged SYK Gibbs states can be approximated to high fidelity from only tens of samples by feeding empirical {p_q, ν_q} into the reconstruction formula.","Nonanalyticities of entanglement entropy at Lifshitz transitions remain visible in the permutation-symmetrized reduced state of a spatial interval.","The same linear data set yields a concrete density matrix on which state-level diagnostics (entropy, fidelity, etc.) can be evaluated without measuring high-order correlators."],"fun_headline_variants":["PI fermion states fixed by counts and one mode","Linear observables suffice for PI fermion tomography","Number stats plus one mode determine PI fermions","Fermion PI tomography needs only p_q and ν_q","Symmetrized fermion states set by sector occupations"],"cache_read_input_tokens":128,"weakest_assumption_plain":"The empty-uniform-mode and occupied-uniform-mode pieces of each fixed-particle-number sector must transform as two inequivalent irreducible representations of the permutation group; if either representation split or they became equivalent, extra free parameters would survive and the two measured numbers per sector would no longer fix the symmetrized state.","fun_headline_variants_meta":{"raw":{"variants":["PI fermion states fixed by counts and one mode","Linear observables suffice for PI fermion tomography","Number stats plus one mode determine PI fermions","Fermion PI tomography needs only p_q and ν_q","Symmetrized fermion states set by sector occupations"]},"model":"grok-4.5","effort":"low","cost_usd":0.003975,"raw_usage":{"total_tokens":1263,"prompt_tokens":792,"num_sources_used":0,"completion_tokens":59,"cost_in_usd_ticks":39748000,"prompt_tokens_details":{"text_tokens":792,"audio_tokens":0,"image_tokens":0,"cached_tokens":256},"completion_tokens_details":{"audio_tokens":0,"reasoning_tokens":412,"accepted_prediction_tokens":0,"rejected_prediction_tokens":0}},"tokens_in":792,"tokens_out":59,"duration_ms":6279,"temperature":1.0,"reasoning_tokens":412,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-07-30T18:20:22.141756+00:00","model_set":{"reader":"grok-4.5"},"falsifier":"Exhibit a U(1)-symmetric fermionic state for which two different permutation-symmetrized density matrices share the same {p_q, ν_q}, or show that the characters of the two exterior-power representations of a transposition agree for some half-filled N=2q, contradicting the End Matter calculation that underpins Schur’s lemma.","supporting_citations":[],"review_version":1}