{"id":"7911e81a-fb49-4c20-94d7-ef9a8fe19517","arxiv_id":"2607.25855","paper_version":1,"verdict":"REJECT","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"high","formal_verification":"none","parameter_count":0,"one_line_summary":"Extends Givon–Hald–Kupferman's existence proof for weak solutions of the orthogonal dynamics equation from stationary Hamiltonian systems to quasicontraction (dissipative) semigroups, with conditional uniqueness for regular solutions.","lead":"This paper proves that the 'orthogonal dynamics' equation inside the Mori–Zwanzig projection formalism has weak solutions for a much wider class of systems than previously known — dissipative, non-Hamiltonian dynamics given by quasicontraction semigroups, going beyond the 2005 Hamiltonian-only proof. The abstract functional-analytic theorems are sound, but the step connecting them to Mori–Zwanzig contains a gap, so the headline claim needs repair.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The reader's rejected bridge lemma is repairable: the numerical-range chain in Lemma III.1 is stated with invalid inclusions, but a closure argument yields the needed bound.","rationale":"The abstract machinery in Sec. II is sound: the energy estimate, Riesz-based existence theorem, growth bound, and uniqueness for regular solutions all check out. The advertised main result, however, is carried by Lemma III.1, and the proof of that lemma contains an invalid numerical-range chain. The paper claims (QLQ)†=QL†Q and Num((QLQ)†)⊆Num(QL†Q), 'by definition of the adjoint' and via 'Num(\\overline{T})⊆Num(T)'. Both claims are false without closure qualifications: the correct identity is (TQ)†= \\overline{QL†Q}, and since the numerical range of a closure can be strictly larger than the numerical range of the original operator, the set inclusion does not hold in general. The reader identified the numerical-range transfer as the weak point, which is exactly right; but the reader's specific claim that D((QLQ)†)⊆D(A†) is opposite to what holds in the natural operator-matrix picture, where D(A†)⊆D((TQ)†). More importantly, the gap is repairable: one can obtain the required bound ω(A†)≤λ_min by noting that (TQ)† is the closure of QL†Q and that sup Re over the closure of a numerical range is the same as over the original set. The rest of the paper—Zwanzig projection density, the damped-oscillator example, and the conditional uniqueness—does not reveal further load-bearing defects. Therefore the central claim is likely mathematically correct, but the written proof of Lemma III.1 needs revision; REJECT is too strong, and CONDITIONAL is the appropriate verdict.","tokens_in":15246,"tokens_out":60327,"duration_ms":576989,"concrete_test":"Independently re-derive Lemma III.1, replacing the asserted chain with: (i) (TQ)† = \\overline{QL†Q} for T = \\overline{QL}; (ii) D(A†)⊆D((TQ)†) and A†=(TQ)†|_X; (iii) the limiting inequality Re(S†y,y)≤λ_min∥y∥². If this derivation succeeds, the numerical-range transfer is repairable and the REJECT should become CONDITIONAL; if the limiting inequality fails for a concrete example (e.g., H=L²(0,1)⊕C² with L=-∂_x on the graph of traces), the lemma is genuinely false.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The load-bearing defect is Lemma III.1's numerical-range transfer: Num(A†)⊆Num((QLQ)†)⊆Num(QL†Q)⊆Num(L†). With T = \\overline{QL} and QLQ understood as TQ, the correct adjoint identity is (TQ)† = \\overline{QT†} = \\overline{QL†Q}, not (QLQ)† = QL†Q. Thus QL†Q is a restriction of (QLQ)†, not an extension; the graph inclusion runs QL†Q⊆(QLQ)†, not the reverse. The asserted set inclusion Num((QLQ)†)⊆Num(QL†Q) is false in general, and the cited 'Num(\\overline{T})⊆Num(T)' is not a valid operator fact. The first inclusion D(A†)⊆D((QLQ)†) actually holds, so the reader's 'opposite nesting' diagnosis is not the right one. The real gap is that the paper has not shown the large adjoint (the closure) inherits the numerical-range bound from QL†Q. This gap is repairable: because sup Re over the closure of a numerical range equals the original sup, a limiting argument from R=QL†Q gives sup Re Num((QLQ)†)≤λ_min, and then ω(A†)≤λ_min. So the central theorem appears true, but the proof as written is invalid.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper proves an abstract weak-solution theorem for the Cauchy problem u' = Au + f on a Hilbert space, under a numerical-range (logarithmic norm) condition on A† (Theorem II.5), with a priori growth bounds and uniqueness for sufficiently regular solutions (Lemma II.6, Corollary II.7). It then applies this to the Mori–Zwanzig orthogonal dynamics by taking A to be the part of (the closure of) QL on X = R(Q) and claiming that the numerical-range condition follows from the quasicontraction property of the semigroup generated by L (Lemma III.1). A density lemma for Zwanzig's projection (Lemma III.3) and a generation lemma for Fokker–Planck-type operators (Lemma III.4) are used to exhibit the damped harmonic oscillator as an example. The abstract section is written carefully, and the limitations (uniqueness gap, regularity) are openly discussed.","tokens_in":15443,"tokens_out":17798,"duration_ms":172158,"significance":"If the application lemma were correct, this would be a meaningful generalization of Givon–Hald–Kupferman: it would replace the Hamiltonian/skew-symmetric hypothesis by quasicontractivity and thus cover non-Hamiltonian infinite-rank projection settings. The abstract existence result in Sec. II is valuable independently: the energy estimate (Lemma II.4) and the Riesz construction (Theorem II.5) appear correct, and the honest discussion of the existence/uniqueness gap is a positive feature. However, the bridge to the Mori–Zwanzig setup (Lemma III.1) contains a load-bearing operator-theoretic error, and the example depends on a proof delegated to the author's earlier paper. The central claim is plausible and likely repairable, but the present text is not correct in the application.","major_comments":[{"comment":"The proof that ω(A†) < ∞ rests on the chain Num(A†) ⊆ Num((QLQ)†) ⊆ Num(QL†Q) ⊆ Num(L†). Two steps are not valid as written. (i) If A is the part of \\overline{QL} in X (as stated at the start of §III), then A extends QLQ|_X, so adjunction gives only D(A†) ⊆ D((QLQ|_X)†); the manuscript's D(A†) ⊆ D((QLQ)†) is not justified, and the text shifts between A = \\overline{QL}|_X and 'the part of QLQ in X'. (ii) If (QLQ)† is identified with \\overline{QL†Q}, the set inclusion Num((QLQ)†) ⊆ Num(QL†Q) is false; the valid statement Num(\\overline{T}) ⊆ \\overline{Num(T)} preserves sup Re but not the set. The conclusion can likely be repaired by a limiting argument using the fact that for x ∈ X, (QL†Qx, x) = (L†x, x), but that argument is absent. Since Lemma III.1 is the only connection between Theorem II.5 and the Mori–Zwanzig equation, the advertised generalization is not proved as written.","section":"§III, Lemma III.1"},{"comment":"Lemma III.4 asserts that the closure of (F·∇, C^1_c) in L^2_ρ generates the quasicontraction semigroup U(t) with U(t)x = x∘φ_t. The proof in Appendix B verifies only the L^2 growth bound on C^1_c and then says that the extension to a semigroup and the generator identification 'suffices to repeat the line of argument in [52, Theorem 4.1]'. Because this lemma is the only source of the quasicontraction property in the worked example, the example is not self-contained and cannot be checked by the reader. Please include a full proof of the generation statement, or state it as a known theorem with a complete proof sketch. The same applies to the identification of the generator L as the closure of (F·∇, C^1_c), which is used to compute L† in §III.B.","section":"Appendix B / Lemma III.4"},{"comment":"The operator A is introduced as 'A := \\overline{QL}|_X' and a few lines later as 'the part of QLQ in X'. These are different operators in general (the paper itself cites such an example in the discussion of QL|_X vs. \\overline{QL}|_X). The proof of Lemma III.1 uses the identification as if the two definitions coincided. This ambiguity is not merely notational: the adjoint inclusions and the numerical-range transfer depend on which operator is meant. Please fix the definition and re-derive Lemma III.1 accordingly.","section":"§III, definition of A"}],"minor_comments":[{"comment":"The notation s_ε is used both for the approximating simple functions and for the mollified function, which is confusing. Also, the statement that s_ε ∈ H^1_0 is correct only because the convolution with a mollifier supported on [−ε,ε] gives zero trace at the endpoints, despite being nonzero just inside the interval; a short explanation would help the reader.","section":"§II, Lemma II.2"},{"comment":"The abstract says 'using the bump function as initial distribution', but ρ is used as the weight in L^2_ρ and in the Zwanzig projection, not as an initial distribution for the dynamics. Rephrase as 'reference density' or 'phase-space density'.","section":"§III.B"},{"comment":"There are several minor typographical issues: 'We only proof the first statement' should be 'We only prove'; 'the proof of the second statement is identical' should be 'similar'; and the introduction uses cP L for the continuous extension of P L without defining the notation.","section":"General"}],"recommendation":"major_revision","confidential_remarks":"The abstract theorem (Sec. II) is sound and likely publishable; the problem is the bridge lemma. If the author repairs the numerical-range argument in Lemma III.1 and supplies a self-contained proof of Lemma III.4, I would be willing to accept a revised version. I do not see grounds for rejection if these are fixed, but the current manuscript is not correct as written."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Dear colleague,\n\nThis paper is worth reading for the abstract part alone. Section II is a clean generalization of the existence proof in Givon–Hald–Kupferman: the energy estimate (Lemma II.4) is a straightforward Gronwall argument, Theorem II.5 is a standard Riesz construction, and Lemmas II.6–II.7 give conditional growth bounds and uniqueness for regular solutions. Moving from GHK's skew-symmetric case (ω=0) to finite logarithmic norm (ω<∞) is genuinely new, and the paper is honest that uniqueness for general weak solutions remains open (Sec. II.C).\n\nThe application to the Mori–Zwanzig setting is where the proof gets shaky. Lemma III.1 claims a numerical-range transfer chain:\n\nNum(A†) ⊆ Num((QLQ)†) ⊆ Num(QL†Q) ⊆ Num(L†).\n\nThe last inclusion is fine, and the first is actually true — D(A†) ⊆ D((QLQ)†) does hold. The middle inclusion Num((QLQ)†) ⊆ Num(QL†Q) is asserted \"by definition of the adjoint\", which it is not. Graph inclusions run the other way: QL†Q is a restriction of (QLQ)†, not an extension, and you cannot pass the numerical range to the closure without an argument. The reader's report gets the nesting direction wrong in places, but the real issue stands: the proof as written does not establish the needed bound on ω(A†). The good news is that the gap is likely repairable. Since only sup Re Num matters, and taking the closure does not change that sup, a limiting argument from R = QL†Q should give sup Re Num((QLQ)†) ≤ λ_min, hence ω(A†) ≤ λ_min. So the central theorem is probably true, but the written proof does not yet prove it.\n\nTwo smaller things. Lemma III.4 delegates the generation proof for the example to the author's own [52] (\"it suffices to repeat the line of argument\"); that is load-bearing for the example and should be spelled out. And the paper itself concedes (Sec. IV) that simultaneous estimates on the logarithmic norms of A and A† are hard — which is exactly where the gap sits.\n\nWho is this for? People working on the foundations of the Mori–Zwanzig formalism, and anyone who wants a clean worked example of Friedrichs-type weak solution theory. It is an honest, mostly well-made paper with a broken bridge that needs repairing. I would send it to a serious referee; with a corrected Lemma III.1 it could be published. In its current form I wouldn't cite the application.","headline":"The abstract existence theorem is solid, but the bridge to the Mori-Zwanzig setting (Lemma III.1) has a real, repairable gap; the paper deserves a serious referee.","tokens_in":16073,"tokens_out":3822,"would_cite":false,"duration_ms":34709,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["47D06","47B44","47A12"],"pacs":[],"model":"deepseek-v4-flash","headline":"The paper proves existence of weak solutions to the orthogonal dynamics equation in the Mori–Zwanzig formalism for nonstationary, non-Hamiltonian systems governed by a quasicontraction semigroup, with growth bounds and uniqueness for regula","keywords":["Mori–Zwanzig projection operator formalism","orthogonal dynamics equation","weak solutions","quasicontraction semigroup","infinite-rank projections","abstract Cauchy problem","generalized Langevin equation","numerical range"],"falsifier":"Compute the numerical range of A† for a small finite-dimensional quasicontraction example with the paper's projection structure; if any point has real part greater than sup Re Num(L†), the containment chain in the paper's Lemma III.1 is false and the existence claim for that class fails.","tokens_in":14911,"feed_emoji":"⚛️","tokens_out":12499,"duration_ms":110230,"temperature":0.7,"pith_summary":"The paper targets a long-standing gap in the Mori–Zwanzig projection-operator technique: the orthogonal dynamics equation, which textbook derivations routinely treat as well-posed, had previously been shown to have weak solutions only for stationary Hamiltonian systems. The author generalizes that result to any evolution given by a strongly continuous quasicontraction semigroup — covering many nonstationary and non-Hamiltonian processes — provided the projection's range contains a dense set of smooth test functions and the generator and its adjoint are densely defined on that range. The proof works directly with weak solutions defined by testing against smooth solutions of the adjoint equation, using an energy estimate controlled by the logarithmic norm of the adjoint operator. This gives the Mori–Zwanzig construction a legitimate mathematical footing for a broad class of dissipative and driven systems.","feed_headline":"Weak solutions exist for Mori–Zwanzig orthogonal dynamics","feed_subtitle":"Extends the existence guarantee to non-Hamiltonian, nonstationary systems governed by quasicontraction semigroups","key_machinery":"The central object is the orthogonal dynamics generator A = \\overline{QL}|_X, the part of the closure of QL in the range of the projection Q. The machine that carries the argument is an adjoint-test-function weak formulation: instead of solving u' = Au + f directly, one tests against smooth functions v satisfying the adjoint equation, defines a Hilbert space via the energy estimate ∥v∥² + ∥v(0)∥² ≤ γ²(∥v(T)∥² + ∥E+v∥²), and obtains u from the Riesz representation theorem. The identity that transfers boundedness from the original generator to the projected problem is the numerical-range containment Num(A†) ⊆ Num((QLQ)†) ⊆ Num(QL†Q) ⊆ Num(L†). The smaller operator QL|X (restricted before closu","core_discovery":"Working in a Hilbert space, the paper defines weak solutions for u' = Au + f, u(0) = g via adjoint test operators. If the numerical range of A† lies in a left half-plane, an energy estimate produces a weak solution by the Riesz representation theorem. In the Mori–Zwanzig setting, with A the part of the closure of QL in X = range(Q), the numerical-range bound transfers from L† to A†, so weak solutions exist whenever the evolution is a quasicontraction semigroup and LQ, L†Q are densely defined. The same estimate gives growth bounds and uniqueness for regular solutions. The conditions are verified for Zwanzig's projection under a C¹ density, with the damped harmonic oscillator as example.","pith_inferences":["A testable next step is to check the numerical-range containment chain in finite-dimensional truncations; a counterexample there would force an extra hypothesis on the projection, since the paper's discussion already notes that simultaneous estimates on A and A† are difficult.","Because uniqueness is only proved for regular solutions, the memory kernel in the generalized Langevin equation derived from the weak solution may not be uniquely pinned down in L²; deciding whether weak solutions are unique in L² would resolve this.","The method needs only a finite logarithmic norm, so the same route should apply to other dissipative generators — e.g. Fokker–Planck operators with confining drift — beyond the damped-oscillator example.","The semigroup-generation lemma behind the damped-oscillator example is delegated to a companion article; if that lemma fails, the illustration loses its support, though the abstract existence theorem would remain intact."],"forward_implications":["The orthogonal dynamics equation has a weak solution for infinite-rank projections in the non-Hamiltonian, nonstationary case, not just stationary Hamiltonian evolutions.","Any orthogonal projection whose range contains a dense set of compactly supported C¹ functions — including Zwanzig's projection with a C¹ density — falls under the theorem.","Regular weak solutions satisfy an explicit exponential growth bound, and two regular solutions with the same data and forcing must coincide.","The damped harmonic oscillator with bump-function density provides a concrete non-Hamiltonian system where the existence and uniqueness claims apply."],"fun_headline_variants":["Weak solutions proven for Mori-Zwanzig with quasicontraction semigroups","Mori-Zwanzig existence proof now covers non-Hamiltonian systems","Nonstationary non-Hamiltonian Mori-Zwanzig gets weak solution guarantee","Orthogonal dynamics weak solutions exist for quasicontraction evolutions","Existence proof for Mori-Zwanzig extends to damped harmonic oscillator"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The existence proof stands on an unproved numerical-range inheritance: that the adjoint of the projected generator has numerical range bounded above by that of the original generator's adjoint.","fun_headline_variants_meta":{"raw":{"variants":["Weak solutions proven for Mori-Zwanzig with quasicontraction semigroups","Mori-Zwanzig existence proof now covers non-Hamiltonian systems","Nonstationary non-Hamiltonian Mori-Zwanzig gets weak solution guarantee","Orthogonal dynamics weak solutions exist for quasicontraction evolutions","Existence proof for Mori-Zwanzig extends to damped harmonic oscillator"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000183,"raw_usage":{"total_tokens":1171,"prompt_tokens":784,"completion_tokens":387,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":528,"completion_tokens_details":{"reasoning_tokens":285}},"tokens_in":528,"tokens_out":387,"duration_ms":4460,"temperature":1.0,"reasoning_tokens":285,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-01T01:22:27.311847+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute the numerical range of A† for a small finite-dimensional quasicontraction example with the paper's projection structure; if any point has real part greater than sup Re Num(L†), the containment chain in the paper's Lemma III.1 is false and the existence claim for that class fails.","supporting_citations":[],"review_version":1}