{"id":"32b38d13-fea9-4574-804d-a54aabe65558","arxiv_id":"2505.13037","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"For the U(1)^3 toy model, the Hamiltonian renormalisation flow built from Narnhofer-Thirring or Fock inputs has fixed points that coincide with the previously known exact continuum solutions, with explicit convergence estimates for local lattice initial data.","lead":"This paper applies the authors' Hamiltonian renormalisation scheme to the U(1)^3 toy model of 4D Euclidean quantum gravity, a self-interacting gauge theory. It claims that, when the finite-resolution inputs mimic the model's known exact solutions, the renormalisation flow lands on those same solutions, and it computes convergence for more local starting data.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Groupoid fixed point rests on an unproven cancellation of perpendicular epsilon-shifts in Eqs. (3.35)-(3.38); the adapted flow may build the answer into the input, so the abstract overclaims.","rationale":"The paper's central claim has two halves. The Fock/algebroid half is largely an application of [17] and [12]; the paper itself flags in Section 4.2 and the conclusions that anomaly-free closure for non-trivial covariance is not checked, so the exact solution is recovered only at the level of quadratic forms. The Narnhofer-Thirring/groupoid half is where the load-bearing gap sits. The fixed point for the state is immediate because the input states are restrictions of the continuum state. The fixed point for the constraints rests on the cancellation of perpendicular ε-shifts, which is asserted but not proved. The toy model in Section 3.1.1 shows that the choice of shifts is not neutral: forward, backward, or symmetric differences would not have the same property. This makes the adapted flow look selected to force the desired fixed point, rather than derived from an independent blocking principle. The explicit convergence estimates in Section 3.2.2 give the paper real content and are not in question; however, they compute the flow of the adapted equations (3.64), whose derivation assumes the same cancellation. The reader's CONDITIONAL verdict is appropriate: accept only if the cancellation is proved, the adapted flow is justified independently, or the claim is restated as a consistency check. Our stress-test identifies the same weakest assumption, so no change to the verdict is needed.","tokens_in":36886,"tokens_out":6185,"duration_ms":66365,"concrete_test":"Symbolically implement, for a small resolution (e.g., M=1 or M=3) and low polynomial degree, the regulated operators (3.35) acting on the orthonormal basis ρ_{3M}(W_{3M}[F_{3M}])Ω, and compute the matrix elements of [−iρ(D^ε_{3M}[u])]^N and [−iρ(H^ε_{3M}[N])]^N between states with F_{3M}, F'_{3M} ∈ L_M for N=1,2,3. Verify the claim that all terms involving nonzero perpendicular shifts kε + lε² P_{M3M}(x, ·) with k+l>0 vanish. If any survive, Eq. (3.38) and the flow equations (3.64) are invalid; if none survive, repeat with the forward-derivative regularisation (shifts ε,0) to check whether the fixed point is regulisation-independent.","verdict_should_be":"UNCHANGED","load_bearing_attack":"In Section 3.1.2 the entire groupoid fixed-point statement for the constraints depends on a cancellation claim introduced after Eq. (3.35): in the N-th power of the regulated constraint, shifts of the form kε + lε² along the perpendicular projector P_{M3M} are asserted to drop out of matrix elements between states labelled by F_M ∈ L_M, leaving only the projected vector field X^{ε,M}. The paper explicitly defers the proof (\"Writing this out in detail is a tedious exercise left to the interested reader\"). If this combinatorial cancellation fails, Eq. (3.38) does not follow and the flow is not fixed; the same mechanism also underlies the discretised flow equations (3.64) and hence the convergence computation of Section 3.2. The choice of strictly positive shifts ε, ε² is essential (see the paragraph after Eq. (3.33)), and no independent argument shows that this adapted flow, rather than the naive flow (3.14), is the correct blocking of the continuum theory of [15]. The central claim is therefore conditional on an unproved and apparently tailored regulisation; the abstract states the conclusion without this caveat.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript applies the Hamiltonian renormalisation scheme developed earlier in this series to the U(1)^3 model of Euclidean quantum gravity in 3+1 dimensions, which is self-interacting. For the Narnhofer–Thirring (groupoid) representation, the authors show that the state family is already fixed under coarse graining because the delta state restricts to itself (3.11)–(3.13), and they introduce an epsilon, epsilon^2-regularised derivative to define a projected flow of the exponentiated constraints. They argue that the perpendicular shifts cancel in matrix elements, so the fixed point is the continuum constraint action of [15]. In the discretised formulation, they derive explicit flow equations for the couplings (3.64) and show, under a compact-momentum-support assumption on the lapse, that local initial data converge geometrically to the projected fixed point (3.75)–(3.92). For the Fock (algebroid) representation with trivial covariance, the blocked quadratic forms coincide with the continuum ones at the zeroth step, and the no-anomaly result of [17] is carried over; for non-trivial translation-invariant covariance, the fixed-point statement is extended, but anomaly-free closure is left to future work.","tokens_in":37071,"tokens_out":7822,"duration_ms":79948,"significance":"If the central claim holds, the paper is a valuable step: it provides a concrete 3+1-dimensional self-interacting model in which Hamiltonian renormalisation has computable fixed points that match known continuum solutions, with explicit estimates and a careful discussion of the discontinuity of Narnhofer–Thirring representations. The toy model in Section 3.1.1 is instructive, the intertwining identity (3.66) is clean, and the convergence estimates in Section 3.2 are explicit and reproducible. The significance is, however, conditional: the groupoid fixed point relies on an unproved combinatorial cancellation, and the adapted flow is not independently justified as the correct blocking procedure. The paper is therefore more convincing as a consistency check of the proposed framework than as an unconditional renormalisation proof.","major_comments":[{"comment":"The groupoid fixed-point statement for the constraints depends on the claim that the perpendicular epsilon- and epsilon^2-shifts proportional to P_{M3M} cancel in matrix elements of powers of the regulated constraint, leaving only the projected vector field X^{epsilon,M}. The paper explicitly defers the proof (\"Writing this out in detail is a tedious exercise left to the interested reader\"), but this cancellation is load-bearing: without it Eq. (3.38) does not follow, and the same mechanism underpins the discretised flow equations (3.64) and the convergence computation of Section 3.2. Since the choice of strictly positive shifts epsilon, epsilon^2 is essential (paragraph after Eq. (3.33)), the correctness of the adapted flow cannot be separated from this unproved combinatorial statement. I ask that a complete proof be supplied, or that the abstract and Section 5 be qualified to state explicitly that the groupoid fixed point is obtained under this unproved cancellation assumption.","section":"§3.1.2, Eqs. (3.35)–(3.38)"},{"comment":"The adapted flow is introduced because the naive flow (3.14) returns zero for all exponentiated constraints. The authors correctly identify that projection and exponentiation do not commute in the Narnhofer–Thirring representation and therefore replace the weak-operator limit by a discrete-topology limit with positive epsilon-shifts. However, no independent argument establishes that this adapted flow, rather than some other regularisation of the non-commuting operations, is the correct blocking of the continuum theory of [15]. As it stands, the construction is tailored so that the projected matrix elements reproduce the input quantisation, which creates a circularity risk for the claim that the flow 'finds' the fixed point rather than being engineered to do so. The paper should either justify the adapted flow from a general principle, such as a locality or continuity axiom, or explicitly reformulate the claim as a consistency check of the framework rather than a predictive renormalisation result.","section":"§3.1.1–§3.1.2"},{"comment":"The convergence proof for the discretised flow assumes compact momentum support of the lapse N (and hence of N,b), and the extension to general N is left to the reader (\"We leave the details to the interested reader\"). The fixed point of (3.64) is exact without this assumption, but the claimed convergence of local initial data to that fixed point is only established under the compact-support hypothesis. Given that Section 3.2 is presented as the 'real space' block-spin analysis, this is a substantial gap; at minimum the abstract and Section 5 should state the condition under which convergence is proven.","section":"§3.2, Eqs. (3.75)–(3.92)"},{"comment":"For non-trivial translation-invariant covariance, Section 4.2 establishes that the blocked couplings coincide with the fixed point and that the flow is fixed at zeroth order, but the paper explicitly states that anomaly-free closure of the constraint algebra has not been checked in this case. Since the algebroid flow is advertised in the introduction to Section 4 as showing 'not only that the quadratic forms do flow to their correct limit but also that the constraint algebra closes without anomalies', the scope of that claim should be restricted to the trivial-covariance case of Section 4.1, or the anomaly check must be supplied. This does not invalidate the central claim involving the exact solutions of [17], but it is necessary for the accuracy of the presentation.","section":"§4.2 and §5"}],"minor_comments":[{"comment":"There appears to be an index inconsistency in H_loc M: the two electric-field factors are written e^a_{M,k}(m)e^b_{M,k}(m), which conflicts with the antisymmetric epsilon^{jkl} and with the corresponding fixed-point expression (3.72). Please correct to the intended antisymmetrised expression e^[a_{M,k}(m)e^b]_{M,l}(m).","section":"§3.2, Eq. (3.50)"},{"comment":"In the toy model, the Weyl elements are introduced as W[F]=e^{-iF^I A_I} and W[F]=e^{-iG_I E^I}; the second should presumably be W[G]=e^{-iG_I E^I}.","section":"§3.1.1, Eq. (3.15)"},{"comment":"The definition of the forward derivative '[partial_{M,b} f_M](m)=M[f_M(m+delta_b)-f_M(m) with the lattice vector with components...' is missing a closing parenthesis; please fix the typo.","section":"§3.2, after Eq. (3.50)"},{"comment":"Section 3.2.3 is only a sketch: the decay assumptions on the Fourier transform are described qualitatively, and the details are left to the reader. Please spell out the concrete decay condition and state the resulting convergence statement, even if briefly.","section":"§3.2.3"},{"comment":"The abstract says 'if one uses as input algebras and states in analogy to those used in the recent exact solutions', while the body and Section 5 say 'using as input algebras and states that were used'. Please align the wording with the actual content, especially given the qualifications identified in Sections 3.1.2 and 3.2.","section":"Abstract and §5"}],"recommendation":"major_revision","confidential_remarks":"The manuscript is a serious contribution and the central idea is plausible, but the groupoid fixed point hinges on an unproved combinatorial cancellation and on the choice of an adapted flow that is not independently justified. I recommend requesting a complete proof of the cancellation and a clearer conceptual justification of the adapted flow in the revision; if the proof cannot be supplied, the abstract should be weakened to a consistency-check statement. The non-trivial-covariance anomaly issue in Section 4.2 should also be flagged explicitly in the main text."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The paper is worth engaging, but the abstract overstates what is shown. The genuinely new content: this is the first self-interacting 3+1-dimensional model treated in this Hamiltonian renormalisation series, and the constraints are nonlinear. The explicit finite-step convergence estimates for local lattice starting points in Section 3.2 are real and reproducible, and the discussion of how Narnhofer–Thirring discontinuity forces a discrete-topology adaptation of the flow is thoughtful. The authors also honestly flag that anomaly-free closure for nontrivial covariance is not checked in the algebroid case.\n\nThe local discretisation flow is the strongest part. The flow equations (3.64) are linear in the couplings, the fixed point is checked directly, and the estimates are uniform in the lattice site, giving exponential convergence under a stated compact-momentum-support assumption. That is a solid, concrete computation. The Fock/trivial-covariance section leans on previous results but the adaptation to projections is clean, and the translation-invariant covariance extension in Section 4.2 is a useful check.\n\nThe main soft spot is the groupoid fixed point. The entire claim in Section 3.1 depends on the assertion that perpendicular epsilon-shifts cancel in powers of the regulated constraint, and the proof is explicitly deferred to \"the interested reader.\" Without that proof, Eq. (3.38) does not follow. More structurally, the adapted flow with strictly positive shifts epsilon and epsilon^2 is motivated by the need to make projected matrix elements survive in the Narnhofer–Thirring representation; no independent argument shows this is the correct blocking of the continuum theory rather than a prescription chosen to reproduce it. So the abstract's \"flow finds as fixed point those exact solution theories\" is too strong. What is actually demonstrated is a consistency check of the method, conditional on a regulation choice that has not been independently justified. The circularityburden is real, but the paper does contain enough detail that the gap is identifiable and potentially fixable.\n\nWho is this for: researchers in canonical quantum gravity renormalisation and constructive QFT who work with the U(1)^3 model or with Hamiltonian block-spin methods more broadly. It deserves a serious referee, because it is substantive and clearly written, and because the central gap is a missing proof rather than a vague hand-wave. I would not cite it as evidence of a theorem unless the cancellation is supplied, but I would cite it as a development if that gap is closed. Recommendation: send to peer review, with a major-revision expectation focused on Section 3.1.","headline":"Genuinely new application of the Hamiltonian renormalisation program to a self-interacting 3+1D toy model, with real convergence estimates but a headline fixed-point claim that is partly built into the input and should be presented as a consistency check.","tokens_in":37642,"tokens_out":1487,"would_cite":false,"duration_ms":17960,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":["04.60.-m"],"model":"deepseek-v4-flash","headline":"The Hamiltonian renormalisation flow of the $U(1)^3$ model of Euclidean quantum gravity, fed with the states and algebras of the known exact solutions, has those exact solutions as its fixed point.","keywords":["Hamiltonian renormalisation","U(1)^3 model","Narnhofer-Thirring representation","Fock representation","coarse graining","Dirichlet kernel","constraint algebra","block-spin transformation"],"falsifier":"Carry out the expansion of the $N$-th power of the regulated Hamiltonian constraint (3.35)-(3.38) for $w=2$ and a generic label $F_M$, collecting all terms after setting the orthogonal component to zero; any surviving term of order $\\epsilon^{-1}$ or $\\epsilon^0$ from the orthogonal directions would contradict the claimed fixed point, since the paper asserts only $k=l=0$ terms survive and leaves the proof to the reader. For the discrete Fock flow, repeat the estimate (3.86) with a lapse $N$ whose Fourier coefficients decay polynomially rather than having compact support and check whether the $r \\to \\infty$ limit still vanishes.","tokens_in":36602,"feed_emoji":"🌀","tokens_out":18915,"duration_ms":170281,"temperature":0.7,"pith_summary":"The paper aims to show that the Hamiltonian renormalisation flow of the $U(1)^3$ toy model of Euclidean quantum gravity in $3+1$ dimensions has the known continuum solutions as fixed points. The model is self-interacting, so this is a test of the coarse-graining method on an interacting gauge theory in four spacetime dimensions. For the Narnhofer-Thirring (groupoid) solution the flow is already fixed under coarse graining via smooth projections; for the Fock (algebroid) solution the blocked normal-ordered constraints coincide with the known continuum quadratic forms. Starting instead from a local lattice discretisation, the couplings converge to the same fixed point, in finitely many steps when the lapse has compact momentum support. If the claim is right, coarse-grained Hamiltonian quantisation of this gravity model is consistent across resolutions and recovers the exactly solved continuum constraint algebra.","feed_headline":"Renormalisation flow hits the exact U(1)^3 gravity solution","feed_subtitle":"Block-spin coarse graining recovers the known continuum constraints in both groupoid and Fock quantisation.","key_machinery":"The load-bearing devices are two. The first is the Dirichlet-kernel coarse graining map: a smooth orthogonal projection $P_M$ with image $L_M$ that commutes with spatial derivatives, together with its discretised version $I_M$, $I_M^\\dagger$ and the intertwining identity $\\partial_M I_{M} = I_{M} \\partial_M$; this identity moves the blocking action onto the couplings in the flow equations. The second, needed only for the groupoid flow, is the $\\epsilon$-$\\epsilon^2$ regulated derivative: replacing the non-existent field $A$ in Narnhofer-Thirring representations by shifts $+\\epsilon$ and $+\\epsilon^2$ in the orthogonal projection directions, so that powers of the orthogonal component drop out of projected matrix elements and projection commutes with exponentiation even though the representation is too discontinuous for weak limits. For the algebroid flow the machine is normal ordering of polynomial constraints plus the mode-cut-off limiting pattern taken from the continuum Fock solution.","core_discovery":"The central claim is that block-spin renormalisation of $U(1)^3$ quantum gravity terminates exactly on the theories that were solved in the continuum in earlier work. With the Narnhofer-Thirring state as input, the projected state family is already fixed, and the exponentiated diffeomorphism and Hamiltonian constraints, regulated by the $\\epsilon$ and $\\epsilon^2$ shift prescription, keep their action inside the projected subspace; the groupoid flow is therefore a fixed point for any density weight, limited only by the non-degeneracy condition $\\det(F) \\neq 0$. With the Fock state, the constraint quadratic forms cut off by Dirichlet projections are already the ones obtained by blocking the continuum theory, and the limiting pattern from the continuum solution makes the algebra close without anomalies. When the flow is started from a local real-space discretisation instead, it is not fixed, but the coupling flow converges to the same quasi-local fixed point; for compact momentum support the convergence is exact after finitely many steps and exponentially fast in the iteration number.","pith_inferences":["A direct check of the $\\epsilon$, $\\epsilon^2$ cancellation for the first few powers of the regulated Hamiltonian constraint, which the paper leaves to the reader, would either confirm the groupoid fixed point or expose residual $\\epsilon^{-1}$ terms.","The same shift regularisation may transfer to the full SU(2) theory if the direction-dependent smearing appropriate for density weight $w=1$ is used, but the non-polynomial dependence would make the flow equations far more complex; the paper itself treats that as future work.","Because the groupoid fixed-point mechanism uses orthonormality of the Narnhofer-Thirring basis, the result is representation-specific: a regular representation with the same coarse graining would not automatically give the same fixed point.","Dropping compact momentum support but keeping rapid Fourier decay should preserve convergence of the local flow, since the paper's own estimate only needs $\\sum_{n_0} |\\hat{N}(n_0)|$ to converge; testing this would extend the result beyond the stated assumption."],"forward_implications":["The $3+1$ $U(1)^3$ model becomes a worked example of Hamiltonian block-spin renormalisation for an interacting gauge theory, not just for free fields or lower-dimensional interactions.","In the Narnhofer-Thirring representation the fixed point exists for arbitrary density weight $w$, because the mechanism does not use the polynomial structure of the Hamiltonian constraint; the only condition is that the smearing $F$ stays non-degenerate.","In the Fock case the fixed point is reached at the zeroth step for trivial covariance, and for translation-invariant non-trivial covariance the same coupling flow equations apply, provided the Hamiltonian constraint is polynomial ($w-2=4k$).","For local lattice initial data, the block-spin flow reaches the fixed point after finitely many iterations at fixed resolution when the lapse has compact momentum support, and the convergence is exponential in the iteration count.","The finite-resolution blocked constraint algebra does not close, but its anomalies vanish in the weak operator topology as $M$ tends to infinity, so the continuum limit returns the closed algebra."],"supporting_citations":[{"why":"supplies the exact continuum groupoid solution, including the Narnhofer-Thirring state and the exponentiated constraints that the flow reproduces as its fixed point.","marker":"[15]"},{"why":"supplies the continuum Fock/algebroid solution, the normal-ordered constraint quadratic forms, and the mode-cut-off limiting pattern used to close the constraint algebra.","marker":"[17]"},{"why":"provides the Hamiltonian renormalisation flow framework for interacting theories and the coupling-flow technique that Section 3.2 applies to the lattice constraints.","marker":"[12]"},{"why":"defines the Dirichlet-kernel and wavelet coarse-graining tools (projection, discretisation, derivative intertwining) on which both flows rely.","marker":"[21]"},{"why":"sets out the Hamiltonian projection scheme and the consistency conditions that motivate searching for fixed points.","marker":"[4]"},{"why":"introduces the Narnhofer-Thirring-type irregular state that defines the groupoid input representation for the flow.","marker":"[16]"}],"fun_headline_variants":["Renormalisation pins down U(1)^3 gravity exact solution","Block-spin flow lands on exact U(1)^3 quantum gravity","U(1)^3 gravity: renormalisation ends at known solution","Fixed point reached: U(1)^3 gravity exact solution"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The groupoid fixed point rests on the new $\\epsilon,\\epsilon^2$ shift regulation of the constraints and on taking coarse-graining limits in the discrete topology; if that regulation is not what genuine blocking of the continuum constraints does, the claimed fixed point is an artifact of the prescription.","fun_headline_variants_meta":{"raw":{"variants":["Renormalisation pins down U(1)^3 gravity exact solution","Block-spin flow lands on exact U(1)^3 quantum gravity","U(1)^3 gravity: renormalisation ends at known solution","Fixed point reached: U(1)^3 gravity exact solution"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000749,"raw_usage":{"total_tokens":3293,"prompt_tokens":858,"completion_tokens":2435,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":474,"completion_tokens_details":{"reasoning_tokens":2359}},"tokens_in":474,"tokens_out":2435,"duration_ms":17706,"temperature":1.0,"reasoning_tokens":2359,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-15T20:21:12.731696+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Carry out the expansion of the $N$-th power of the regulated Hamiltonian constraint (3.35)-(3.38) for $w=2$ and a generic label $F_M$, collecting all terms after setting the orthogonal component to zero; any surviving term of order $\\epsilon^{-1}$ or $\\epsilon^0$ from the orthogonal directions would contradict the claimed fixed point, since the paper asserts only $k=l=0$ terms survive and leaves the proof to the reader. For the discrete Fock flow, repeat the estimate (3.86) with a lapse $N$ whose Fourier coefficients decay polynomially rather than having compact support and check whether the $r \\to \\infty$ limit still vanishes.","supporting_citations":[{"cited_title":"Rodriguez Zarate, T","cited_arxiv_id":null,"evidence_quote":"provides the Hamiltonian renormalisation flow framework for interacting theories and the coupling-flow technique that Section 3.2 applies to the lattice constraints."},{"cited_title":"Renormalisation, wavelets and the Dirichlet-Shannon kernels","cited_arxiv_id":"2207.08294","evidence_quote":"defines the Dirichlet-kernel and wavelet coarse-graining tools (projection, discretisation, derivative intertwining) on which both flows rely."},{"cited_title":"Narnhofer, W.E","cited_arxiv_id":null,"evidence_quote":"introduces the Narnhofer-Thirring-type irregular state that defines the groupoid input representation for the flow."}],"review_version":1}