{"id":"39da2945-ee17-46cb-af80-55790535ae63","arxiv_id":"2607.09617","paper_version":2,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":6.0,"correctness_risk":"low","formal_verification":"none","parameter_count":1,"one_line_summary":"Admissible skew-gradient embeddings form an affine gauge family; least-squares and Jacobi criteria produce structure-preserving, decoupled discretizations for NS and CHNS.","lead":"The paper generalizes skew-gradient embeddings for GENERIC systems into an affine family of gauges, with least-squares, regularized, and invariant-preserving selections. This yields structure-preserving, often-decoupled time schemes for Navier–Stokes and Cahn–Hilliard–Navier–Stokes.","discovery_kind":"extension","skeptic_critique":{"model":"grok-4.5","headline":"No significant objection identified beyond the reader's already-flagged compatibility hypothesis.","rationale":"The reader’s strongest claim accurately captures the paper’s algebraic contribution, and the weakest assumption correctly isolates the only non-self-contained step (spatial compatibility). My re-examination of Theorems 3.1–3.2, Propositions 3.4–3.5 and 3.7, and the MAC/CHNS arguments in §4 finds no additional load-bearing gap: the finite-dimensional statements are elementary linear algebra and do not rely on unstated analytic hypotheses. The absence of new numerical experiments is a presentational limitation already noted by the reader, not a flaw in the central mathematical claim. Consequently the CONDITIONAL verdict and its rationale stand without adjustment.","tokens_in":15483,"tokens_out":498,"duration_ms":5174,"concrete_test":"Independently recompute the two scalar coefficients λ_A^{n+1}, λ_J^{n+1} of the regularized GSGE–BDF2 scheme (Eqs. 31–32) on a single MAC time step of the CHNS system with a known exact free-energy-neutral residual; verify that the discrete pairing (R_u,u)+(R_ϕ,μ̄) vanishes to machine precision and that the telescoping free-energy identity of Thm. 4.6 holds. If either fails, the compatibility hypothesis is violated for that discretization.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The algebraic core of the strongest claim is self-contained and holds: admissible gauges form an affine space (Prop. 2.3), the unique minimum-Hilbert–Schmidt representative is given by the weighted wedge (Thm. 3.1), regularization and residual projection follow by the same least-squares argument (Thm. 3.2, Prop. 3.4), and the invariant-preserving construction is a direct Gram projection (Prop. 3.5). The rank-two Jacobi criterion (Prop. 3.7) is standard and correctly applied. The only place the central claim could fail is the inheritance of free-energy neutrality and the discrete Lie-bracket identity under the MAC / periodic SBP discretizations used in §4; that is precisely the weakest assumption already identified by the reader. No deeper internal inconsistency or hidden analytic gap appears in the finite-dimensional theory.","agreement_with_reader":"agree"},"referee_report":{"model":"grok-4.5","summary":"The paper generalizes the authors’ earlier skew-gradient embedding (SGE) by characterizing all two-forms that embed the reversible GENERIC field L∇E (or L∇F) as an affine family of admissible gauges (GSGE). For any positive-definite metric A, weighted least squares selects a unique minimum-Hilbert–Schmidt representative that recovers SGE when A = I; the same construction yields regularized, residual-correcting, and invariant-preserving gauges that retain the entropy or free-energy law. A finite-dimensional rank-two Jacobi criterion is given for the isothermal case. The theory is applied to a compatible MAC discretization of incompressible Navier–Stokes, producing a semi-discrete rank-two Poisson structure that the implicit midpoint rule preserves fully discretely with an exact free-energy law, and to a regularized GSGE–BDF2 scheme for Cahn–Hilliard–Navier–Stokes that preserves mass, dissipates free energy unconditionally, and decouples through two scalar coefficients.","tokens_in":15697,"tokens_out":1208,"duration_ms":9199,"significance":"If the claims hold, the work supplies a clean algebraic toolkit for structure-preserving time discretizations of thermodynamically consistent multiphysics systems: an affine gauge space, a unique least-squares selection principle, regularization at vanishing force, and invariant-preserving projections, all retaining the discrete energy/entropy law while often decoupling subproblems. The finite-dimensional proofs (affine space, uniqueness of ω_A,r, regularized formula, Gram projection, rank-two Jacobi criterion) are self-contained and machine-checkable in principle. The MAC + midpoint construction yields a fully discrete isothermal GENERIC scheme with an exact free-energy identity, and the CHNS scheme inherits mass conservation and unconditional free-energy dissipation from the same gauge algebra. These are concrete, reusable contributions for geometric numerical integration of GENERIC and free-energy systems.","major_comments":[{"comment":"Theorems 4.2–4.3 and 4.6 rest on the hypothesis that the chosen spatial operators (MAC for NS; periodic summation-by-parts for CHNS) exactly inherit free-energy neutrality ⟨∇F, J⟩ = 0 and the algebraic identities needed for the discrete Lie bracket and energy telescoping. The manuscript states the needed identities (skewness of C_h, quadratic homogeneity of J_h, (J_h(u),u)_h = 0) but does not verify them for the concrete centered MAC averages of N_h and K_h, nor does it supply a short appendix or reference that does so for the precise stencil used. Without that verification the fully discrete GENERIC claim and exact energy law remain conditional on an unproved compatibility assumption.","section":null},{"comment":"Section 4.2 and Remark 4.4 introduce a three-parameter regularized differential gauge (ℓ1, ℓ2, ℓ3) and assert second-order consistency when ℓ3 \tau^{2} = O(\tau^{2}) and d^{n+1} > 0, yet no truncation-error analysis or numerical confirmation is given. Because the paper explicitly declines new numerical experiments, a short consistency argument (or a pointer to the order analysis of the recovered SGE–SBDF2 limit) is needed to substantiate that the free parameters do not degrade the formal order of the BDF2 scheme.","section":null}],"minor_comments":[{"comment":"The abstract and introduction cite GuWangSGE2025 / arXiv:2509.18601 as the SGE baseline; ensure the published or final arXiv version is used consistently and that the present paper is self-contained for readers who have not seen that preprint.","section":null},{"comment":"Notation for the musical maps ♭/♯ and the two wedge products (forms vs. maps) is introduced carefully in §2.1 but then used interchangeably; a one-line reminder when switching from ω to the map Y ∧ Z would help.","section":null},{"comment":"Proposition 3.8 asserts local existence of a commuting Y with prescribed invariants; a brief remark on whether the construction extends globally on the affine phase space of CHNS (fixed mass, divergence-free) would clarify the scope of the Poisson reconstruction.","section":null},{"comment":"In (30) the chemical-potential extrapolation uses a convex-splitting-style χ; a short sentence relating this choice to the energy estimate of Theorem 4.6 would make the free-energy telescoping easier to follow.","section":null},{"comment":"Typos: “GuWangSGE2025” formatting in the abstract; occasional missing spaces after commas in multi-line displays; “Theorem 2.2” in Remark 2.4 should be Definition 2.2.","section":null}],"recommendation":"minor_revision","confidential_remarks":"The algebraic core is solid and the paper is a natural, well-executed generalization of the authors’ own SGE work. The two major points are fixable with a short appendix or added paragraphs; I would not escalate to major revision if the authors supply the MAC identity verification and a consistency remark. Fit for a numerical-analysis / geometric-integration journal is good."},"author_rebuttal":null,"desk_editor":{"model":"grok-4.5","letter":"This is a clean, useful extension of the authors' own SGE paper. The real novelty is the affine characterization of admissible two-forms that embed the reversible field, plus the operator-weighted least-squares principle that picks a unique minimum-Hilbert–Schmidt gauge for any positive-definite metric A (recovering SGE when A = I). From the same construction they get regularized gauges, residual corrections, and invariant-preserving projections. The rank-two Jacobi criterion is standard but correctly applied, and they verify it on a MAC discretization of incompressible NS so that midpoint time-stepping yields a fully discrete rank-two GENERIC scheme with an exact free-energy law. For CHNS they give a three-parameter regularized GSGE–BDF2 scheme that preserves mass, dissipates free energy unconditionally, and still decouples through two scalars.\n\nThe finite-dimensional algebra is solid: Prop. 2.3, Thms. 3.1–3.2, Prop. 3.5, and Prop. 3.7 are short and self-contained. Self-citations to their prior SGE and BDF2 papers are legitimate baselines, not circular. The soft spot is exactly the one the reader flagged: both application theorems rest on the hypothesis that the spatial discretization (MAC or periodic SBP) inherits free-energy neutrality and the algebraic identities needed for the discrete Jacobi identity and energy telescoping. That is standard and reasonable, but it is assumed rather than re-proved, and the paper supplies no new numerical tests of the GSGE schemes. The free parameters ℓ1, ℓ2, ℓ3 are explicit and controlled; they do not hide fitting.\n\nThis is for people who already work on structure-preserving integrators for GENERIC/Onsager systems and want a systematic way to choose the skew embedding. It is not a breakthrough in continuum modeling, but it is honest, formally grounded work that a serious referee should see. I would send it out; the compatibility hypothesis and the missing numerics are revision items, not desk-reject items.","headline":"Clean algebraic extension of the authors' SGE work: affine gauges, least-squares selection, and two solid discrete GENERIC schemes, with the usual compatibility caveat and no new numerics.","tokens_in":16311,"tokens_out":517,"would_cite":true,"duration_ms":5430,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["65M06","65P10","37N10","76D05"],"pacs":[],"model":"grok-4.5","headline":"The reversible part of a thermodynamically consistent system can be written many ways; least-squares picks a unique gauge that keeps the energy law and often decouples the numerics.","keywords":["GENERIC","skew-gradient embedding","structure-preserving discretization","Poisson structure","Cahn–Hilliard–Navier–Stokes","energy-stable schemes","gauge freedom"],"falsifier":"On a uniform MAC grid for incompressible Navier–Stokes, compute the discrete Lie bracket of the constructed rank-two fields and check whether it remains inside their span; any nonzero component outside that span falsifies the claimed Poisson structure and the exact discrete free-energy law of the midpoint scheme.","tokens_in":16348,"feed_emoji":"⚖️","tokens_out":722,"duration_ms":5683,"temperature":0.7,"pith_summary":"Thermodynamically consistent continuum models split into reversible and irreversible pieces that must respect energy and entropy balance. Earlier work rewrote the reversible piece by embedding it in a single rank-two skew operator driven by the thermodynamic force, so that the whole system becomes a generalized gradient flow. This paper shows that embedding is not unique: every admissible two-form differs from a reference form by something that still annihilates the force, forming an affine family called generalized skew-gradient embeddings. For any positive-definite metric the unique minimum-norm representative is given by a simple weighted wedge product; the ordinary metric recovers the original embedding. The same formula supplies regularized gauges when the force vanishes, corrections of slightly non-neutral residuals, and gauges that preserve any prescribed list of invariants. A rank-two Jacobi identity then tells when the gauge is Poisson, allowing fully discrete GENERIC schemes. Applied to a marker-and-cell discretization of Navier–Stokes and to a regularized BDF2 scheme for Cahn–Hilliard–Navier–Stokes, the construction yields exact discrete energy laws and inexpensive decoupled solvers.","feed_headline":"One formula picks the unique gauge that keeps energy laws","feed_subtitle":"Least-squares gauges turn reversible physics into gradient flows that stay structure-preserving and often decouple","key_machinery":"The unified least-squares gauge (Theorem 3.1): among all two-forms whose contraction with the thermodynamic force recovers the reversible field, the weighted wedge ω_A,r is the unique minimizer of the A-Frobenius norm, and its action remains matrix-free.","core_discovery":"The admissible two-forms that embed the reversible field form an affine space. For any positive-definite metric A the unique minimum-Hilbert–Schmidt gauge is the weighted wedge (A X)^♭ ∧ r / ⟨A X, X⟩; the identity metric recovers the original skew-gradient embedding, while the same least-squares principle produces regularized, residual-correcting and invariant-preserving gauges that still obey the entropy or free-energy law.","pith_inferences":[],"forward_implications":[],"fun_headline_variants":["Least-squares picks unique gauge for energy-preserving embeddings","Affine space of gauges turns reversible physics into gradient flows","Min-Hilbert-Schmidt gauge recovers SGE and unlocks decoupling","Rank-two GSGE yields discrete GENERIC structure for Navier-Stokes","Regularized GSGE-BDF2 decouples CHNS while dissipating energy"],"cache_read_input_tokens":128,"weakest_assumption_plain":"The spatial discretization must exactly inherit the continuous free-energy neutrality and the algebraic identities that make the discrete Jacobi identity and energy telescoping hold; if those identities are lost, the fully discrete structure and exact energy law collapse.","fun_headline_variants_meta":{"raw":{"variants":["Least-squares picks unique gauge for energy-preserving embeddings","Affine space of gauges turns reversible physics into gradient flows","Min-Hilbert-Schmidt gauge recovers SGE and unlocks decoupling","Rank-two GSGE yields discrete GENERIC structure for Navier-Stokes","Regularized GSGE-BDF2 decouples CHNS while dissipating energy"]},"model":"grok-4.5","effort":"low","cost_usd":0.00422,"raw_usage":{"total_tokens":1288,"prompt_tokens":829,"num_sources_used":0,"completion_tokens":73,"cost_in_usd_ticks":42200000,"prompt_tokens_details":{"text_tokens":829,"audio_tokens":0,"image_tokens":0,"cached_tokens":128},"completion_tokens_details":{"audio_tokens":0,"reasoning_tokens":386,"accepted_prediction_tokens":0,"rejected_prediction_tokens":0}},"tokens_in":829,"tokens_out":73,"duration_ms":4079,"temperature":1.0,"reasoning_tokens":386,"cache_read_input_tokens":128,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-07-14T15:07:39.356107+00:00","model_set":{"reader":"grok-4.5"},"falsifier":"On a uniform MAC grid for incompressible Navier–Stokes, compute the discrete Lie bracket of the constructed rank-two fields and check whether it remains inside their span; any nonzero component outside that span falsifies the claimed Poisson structure and the exact discrete free-energy law of the midpoint scheme.","supporting_citations":[],"review_version":2}