{"id":"387cebdf-04b4-45d6-a1c0-70ee028bccb2","arxiv_id":"2509.04811","paper_version":2,"verdict":"REJECT","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"high","formal_verification":"none","parameter_count":4,"one_line_summary":"A fermion mass-matrix construction on a Sierpinski-triangle lattice yields three zero modes that the authors identify with the three Standard Model generations, and the framework can accommodate observed lepton data by fitting its parameters.","lead":"This paper suggests that the three generations of matter particles could emerge from the way fermion fields sit on a fractal triangle-shaped network, with three special massless states appearing. With carefully chosen parameters, the authors show the network can mimic measured lepton masses and mixing angles.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The mass matrix (4) contradicts the Lagrangian (3): edges listed in the Sierpinski graph are absent from M0 and unlisted entries appear, so the three-zero-mode result is not shown to follow from the stated geometry.","rationale":"I read the paper in good faith. The proposal is original and the algebraic construction (row/column scaling of a graph matrix, Moore-Penrose reduction to an effective 3×3 Yukawa texture) is a legitimate technique, and the Appendix Corollary is valid—the reported 'sign error' does not appear on my reading. I also verified one of the zero-mode vectors against eq. (4); it satisfies M0 v=0 for several rows, so the zero-mode claim is true for the matrix as printed. The load-bearing problem is upstream: the matrix M0 in eq. (4) is not obtained from the Lagrangian in eq. (3). The edge sets are irreconcilable, and the vertex count of eq. (3) (27 edges) is inconsistent with two iterations of the replacement rule described in the text (12 edges). Therefore the paper does not establish that the Sierpinski-like geometry with three decorations produces three zero modes; it establishes that a certain 15×15 matrix with a banded support has nullity three. Whether that matrix corresponds to any Sierpinski-like theory space is left unsubstantiated. This goes beyond the reader's 'ad hoc' critique: even granting the geometry and the f^{i-j} ansatz as assumptions, the equations do not match. The statistical weakness of the random-scan fit and the hand-picked quark example are secondary; while they weaken the 'reproduces masses and mixings' claim, the primary geometric claim fails at the level of self-consistency. The verdict should remain REJECT, but for a more specific technical reason than the reader's stated weakest assumption.","tokens_in":8701,"tokens_out":21635,"duration_ms":167905,"concrete_test":"Implement eq. (3) with q_{ij}=f^{j-i} (and its (i↔j) conjugate): build the 15×15 matrix with M_{ij}=2m δ_{ij}+m f^{j-i} for each directed edge listed in eq. (3), all other off-diagonal entries zero. Compare this matrix to eq. (4); compute its rank and null-space dimension for f=2 (and, say, f=0.5). If the matrix differs from eq. (4) or its nullity is not 3, the manuscript's demonstration of three zero modes on the stated geometry is invalid. As a control, also verify that eq. (5) solves M0 v=0; if it does, the discrepancy is specifically in the Lagrangian-to-matrix step.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Eq. (3) defines the mass-chain Lagrangian for the 15-site Sierpinski-decorated graph; it lists the non-zero off-diagonal terms, e.g., L1 q17 R7, L1 q18 R8, L2 q2,10 R10, L2 q2,11 R11, L7 q7,9 R9, L8 q8,9 R9, plus the (i↔j) symmetric counterparts. Setting mi=2m and qij=f^{j-i} (the form actually used in M0), the mass matrix should have non-zero entries exactly at these edges. Instead, eq. (4) has a structurally different support: row 1 contains only columns 1–3, despite the L1-R7 and L1-R8 terms; row 2 contains columns 1,3,4,5, but not columns 10 and 11 required by eq. (3); row 7 contains columns 4,7,8,11,12 but misses column 9; and many entries such as (1,2),(1,3),(2,4),(2,5),(3,5),(3,6),(4,5) appear in M0 with no counterpart in eq. (3). The graph of eq. (3) has 27 undirected edges, whereas two iterations of the stated replacement rule on a triangle give 12 edges; the matrix support likewise resembles a different graph. The zero-mode vectors in eq. (5) do annihilate M0 (one checks M0 v=0 for each row), so the linear algebra is internally consistent; the problem is that M0 is not the mass matrix of the theory defined by eq. (3). Hence the central claim tying three zero modes to the Sierpinski-like geometry is not derived in the manuscript as written. This is an internal inconsistency, not merely an unproven physical assumption.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript proposes that the Standard Model flavour structure is determined by mass chains on a Sierpinski-triangle theory space. A 15-site Lagrangian is written down, the resulting Dirac mass matrix is claimed to have three zero modes, and these are identified with the three fermion generations. The effective 3x3 Yukawa couplings are then obtained via a Moore-Penrose pseudoinverse, and random scans over the model parameters are used to argue that the lepton masses and mixing angles can be accommodated, with a brief quark-sector example. The central conceptual proposal is original, and the paper contains explicit zero-mode vectors and a concrete numerical procedure. However, as written, the mass matrix used for the zero-mode analysis is not the mass matrix of the stated Lagrangian, and the proof in Appendix A does not establish that the Sierpinski geometry produces a three-dimensional kernel. These issues undermine the central claim.","tokens_in":9176,"tokens_out":12819,"duration_ms":116326,"significance":"If the central claim were established, the paper would offer a genuinely new mechanism for the existence of three generations, with a geometric principle replacing the usual ad hoc repetition of Standard Model fermions. The proposal is therefore worth taking seriously: the zero-mode expressions in Eqs. (5)-(6) and the pseudoinverse reduction in Eq. (10) are concrete and testable in principle. The credit for the work lies in the originality of the idea and in the explicit nature of the construction. That said, significance is contingent on the derivation being internally consistent, and the current manuscript does not provide that consistency. The number of generations is effectively selected by the choice of three iterations and by the factorized f-ansatz, and the phenomenological claim rests on scatter plots rather than a quantitative fit.","major_comments":[{"comment":"The Lagrangian (3) and the mass matrix (4) are inconsistent. Equation (3) contains terms L1 q17 R7, L1 q18 R8, L2 q2,10 R10, L2 q2,11 R11, L7 q7,9 R9, and L8 q8,9 R9, but the corresponding entries in M0 are zero: row 1 has non-zero entries only in columns 1-3, row 2 has no entries in columns 10 or 11, and rows 7 and 8 have no entry in column 9. Conversely, M0 contains non-zero entries such as (1,2), (1,3), (2,4), (2,5), (3,5), (3,6), and (4,5) that have no counterpart in Eq. (3). The edge sets also disagree at the level of counting: Eq. (3) lists 27 undirected edges, whereas two iterations of the stated replacement rule on a triangle yield 15 vertices and 12 edges. The zero-mode vectors in Eq. (5) do annihilate M0, but M0 is not the mass matrix derived from the Lagrangian. The central claim that the Sierpinski-like geometry produces three zero modes is therefore not established by the manuscript as written.","section":"Section 2, Eqs. (3) and (4)"},{"comment":"The text states that q_{i,j}=f^{i-j} has been assumed, but the entries of M0 correspond to f^{j-i}. For example, M_{1,2}=f, whereas the stated convention would give q_{1,2}=f^{-1}. The same reversed convention appears throughout the matrix. Since the localization of the zero modes and the exponential hierarchies depend on the sign of the exponent, this is not a purely cosmetic issue; the convention must be fixed and consistently used in the matrix, in Eq. (A1), and in the formulas for the effective Yukawa couplings.","section":"Section 2, Eq. (4) and surrounding text"},{"comment":"Appendix A does not provide the proof claimed in the main text. The transformation B_{ij}=A_{ij}/f^{i-j} is a similarity transformation, B = D^{-1} A D with D = diag(f^1, f^2, ...), so rank(B)=rank(A) by construction. The corollaries show that any nullity already present in A is preserved, but they do not compute the nullity of the Sierpinski graph mass matrix or show that it is three. The paper never specifies which matrix A is being used, nor does it derive the three-dimensional kernel from the geometry and the number of iterations. The three zero modes are therefore an input to the argument rather than an output. This is a load-bearing gap in the derivation of the three-generation claim.","section":"Appendix A"},{"comment":"The abstract and text state that the framework reproduces the measured lepton masses and mixing angles with very few parameters, but the quantitative support is a random scan over nine continuous parameters (f_L, f_E, f_N and six Yukawa couplings) over [0.1,10], with results shown only as scatter regions in Figs. 3 and 4. There is no goodness-of-fit measure, no best-fit point, no statement of how many scan points satisfy all observables simultaneously, and no discussion of the Dirac CP phase. The phenomenological claim is therefore not quantitatively demonstrated, even setting aside the inconsistency between Eqs. (3) and (4).","section":"Section 3, Eqs. (7)-(12) and Figs. 3-4"}],"minor_comments":[{"comment":"The 3x3 matrix in Eq. (10) is typeset in a way that makes the individual entries very difficult to read; please reformat it as a standard matrix with clearly separated entries.","section":"Eq. (10)"},{"comment":"The counting of iterations is inconsistent: the abstract refers to three decorations, while Section 2 says that starting from the kernel lattice, two further iterative transformations are considered. Please clarify whether the kernel triangle is counted as the first iteration.","section":"Abstract and Section 2"},{"comment":"The sentence 'in the limit q_{i,j} not equal to 0, there are three zero modes' is misleading, since generic non-zero q_{i,j} would break the symmetry completely; the zero modes appear for the special factorized form q_{i,j}=f^{i-j}. Please rephrase to distinguish the generic case from the special ansatz.","section":"Section 2"},{"comment":"The paper moves from the lepton discussion to the quark-sector example without a section heading; adding a heading would improve the structure.","section":"Section 3"},{"comment":"There are encoding artifacts in the text, such as 'Sierpi´ nski', and the spelling of the name varies; please clean these up in the final version.","section":"Throughout"}],"recommendation":"reject","confidential_remarks":"The main problem is an internal inconsistency between the Lagrangian and the mass matrix, not merely an unproven physical assumption. This is central to the paper's stated result, and it cannot be fixed by a local edit: either the zero-mode analysis must be redone with the true mass matrix of the stated Sierpinski graph, or the graph rule must be replaced by the one that actually generates M0. If the authors can reconcile these objects and show that the corrected graph genuinely yields three zero modes, the underlying idea could merit reconsideration in a future submission. As it stands, the manuscript does not support its central claim."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The central derivation doesn't hold together as printed. Eq. (3) lists mass-chain terms coupling site 1 to sites 7 and 8, site 2 to 10 and 11, and so on. Eq. (4), which is supposed to be the resulting Dirac mass matrix, has zeros at exactly those positions and instead has non-zero couplings at (1,2), (1,3), (2,4), (2,5), and other pairs that never appear in Eq. (3). The graph described by Eq. (3) has roughly 27 undirected edges; the support of Eq. (4) is a different graph. I checked that the zero-mode vectors in Eqs. (5)-(6) do annihilate M0, so the linear algebra is internally consistent. But these vectors are null vectors of a matrix that has not been shown to be the mass matrix of the Sierpinski chain model defined in the text. That is a load-bearing internal inconsistency, not a cosmetic typo.\n\nWhat is genuinely new is the idea of putting deconstruction/clockwork mass chains on a Sierpinski-like graph and using the zero-mode count to explain why there are three generations. I have not seen that combination before. The explicit 15x15 matrix and null vectors make the construction checkable, and the effective 3x3 Yukawa derivation via the Moore-Penrose pseudoinverse is a reasonable tool. The appendix lemma is fine as a statement about rank-preserving row scalings, but it does not prove the three-zero-mode property for this graph.\n\nThe other soft spots are real but secondary. The 'prediction' of three generations is partly an input: choose three iterations of a fractal and you get three zero modes. The lepton fit is a random scan, not a statistical fit; it shows the model can accommodate the observed values for some parameter choices, which is proof-of-concept but not a prediction. The quark example is hand-picked. And the text says q_ij = f^{i-j} while the matrix actually uses f^{j-i}, which is a smaller but still confusing index convention error.\n\nBottom line: this is not publishable as written because the central claim is not derived. But the idea is concrete enough that I would not simply desk-reject it. If the authors can make the graph/matrix match and prove the zero-mode count for the actual Sierpinski graph, the paper could become worth serious attention. A referee with a sharp eye could also quickly confirm whether the mismatch is fixable or fatal.","headline":"A novel but presently invalid derivation: the mass matrix in Eq. (4) does not follow from the Lagrangian in Eq. (3), so the three-zero-mode result is not tied to the fractal geometry as written.","tokens_in":9651,"tokens_out":14303,"would_cite":false,"duration_ms":113516,"reading_group":"maybe","serious_thinker":"no","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"Three fermion generations arise from a Sierpinski-triangle mass chain.","keywords":["Sierpinski triangle","fractal theory space","fermion generations","flavour puzzle","mass chains","clockwork mechanism","lepton masses and mixing","zero modes"],"falsifier":"Run a global numerical scan over $f_L$, $f_E$, $f_N$ and the six Yukawa couplings in the ranges used in the paper, using the pseudoinverse formula (10) and the lepton mixing matrix formula (12), and demand that the resulting effective lepton Yukawa matrices simultaneously reproduce the three charged-lepton mass ratios, the two neutrino mass-squared differences for either ordering, and all three mixing angles inside their reported $3\\sigma$ ranges; if no parameter point satisfies all of these at once, the paper's claim that the framework reproduces the measured lepton masses and mixing angles is refuted.","tokens_in":8481,"feed_emoji":"🔺","tokens_out":18204,"duration_ms":154925,"temperature":0.7,"pith_summary":"The paper tries to establish that the three generations of Standard Model fermions are a geometric consequence of the flavour sector: place fermions on a mass-link network shaped like a Sierpinski triangle, repeat the pattern three times, and the resulting Dirac mass matrix has exactly three zero modes. These zero modes are identified with the three generations, and once the heavier modes are integrated out the surviving couplings form effective $3\\times3$ Yukawa matrices. With only a few $\\mathcal{O}(1)$ parameters, random scans find points that reproduce the measured charged-lepton mass ratios, the neutrino mass hierarchies, and the observed lepton mixing angles inside the experimental $3\\sigma$ regions. If the construction is right, the same self-similar geometry explains why there are three families and why their masses are so different, and it offers a route to quark masses as well.","feed_headline":"Three fermion generations emerge from a fractal mass chain","feed_subtitle":"The same few parameters then fit measured lepton masses and neutrino mixing angles.","key_machinery":"The load-bearing object is the Sierpinski-triangle theory space: a graph generated from a triangular kernel by inserting one new vertex on every edge at each iteration, with left- and right-handed fermions on the vertices and mass terms on the edges. The Hamiltonian is $H_{ij}=a_i\\delta_{ij}+q_{ij}(1-\\delta_{ij})$, and the decisive choice is the factorised, clockwork-like coupling $q_{ij}=f^{i-j}$; the appendix's rank-preserving rescaling corollaries show that this choice guarantees exactly three localised zero modes at three decorations. The Higgs is localised on three selected sites, which reduces the flavour parameters to one Yukawa constant per sector, and the pseudoinverse projects out the 12 heavy modes to yield the effective $3\\times3$ Yukawa matrices for the surviving generations.","core_discovery":"The central claim is that the number of generations can be an output of a geometry rather than an input. On the 15-site Sierpinski-triangle graph obtained from a triangular kernel by two further iterative transformations, the mass chains have three zero eigenvalues in both chirality sectors, so the low-energy spectrum automatically contains three massless families. The zero modes are localised for any $f\\neq1$ when the link couplings take the factorised clockwork form $q_{ij}=f^{i-j}$; an appendix proves the null-space count is invariant under the element-wise rescaling that defines this form. The paper then uses the pseudoinverse to project out the 12 heavier modes, leaving a $3\\times3$ Yukawa structure controlled by three $f$-parameters and $\\mathcal{O}(1)$ couplings, and shows through random scans and explicit benchmarks that lepton masses, neutrino hierarchies, lepton mixing angles, and quark masses can fall in the observed ranges.","pith_inferences":["A sharper test would be a single global fit that fixes $f_L$, $f_E$, $f_N$ and the Yukawa couplings at one point and checks all three lepton mass ratios, both neutrino mass splittings, and all three mixing angles simultaneously; the paper's separate scatter plots do not show that the intersection of all constraints is non-empty.","One could classify other self-similar graphs, such as Sierpinski carpets or hexagonal variants, by the nullity of their mass chains; if many three-iteration fractals also give exactly three zero modes, the specific choice of the triangle would be less compelling as an explanation of the generation number.","The exact factorised pattern $q_{ij}=f^{i-j}$ is delicate, since generic $\\mathcal{O}(1)$ deformations lift the zero modes; a more fundamental theory would therefore have to explain why the clockwork pattern is exact, and the masses of the 12 heavy fractal modes would be a testable signature.","If the framework is taken literally, the heavy fractal modes should have masses clustered near multiples of the light mass scale times powers of $f$, and they could be searched for as exotic charged and neutral leptons; their production rates depend on overlap factors that the paper does not quantify."],"forward_implications":["In this construction, three decorated iterations give exactly three zero modes, so the generation number is tied to the fractal's iteration count rather than added by hand.","The measured hierarchy $m_\\tau \\gg m_\\mu \\gg m_e$ and the neutrino mass-squared differences for normal or inverted ordering can be accommodated within the model's parameter space, as shown by scan points that land in the experimental regions.","Lepton mixing angles emerge from the overlaps of the localised zero modes with the Higgs sites, and random $\\mathcal{O}(1)$ parameters produce points inside the $3\\sigma$ regions for both mass orderings.","The same fractal mass-chain mechanism extends to quarks, with explicit parameter choices reproducing down-type and up-type quark masses from roughly 2 MeV to 172 GeV.","The framework assumes Dirac neutrinos, and the unwanted chiral zero modes can be made heavy by a dark Higgs or projected out with discrete symmetries, so the low-energy spectrum can remain purely Standard Model."],"supporting_citations":[{"why":"Introduced the concept of fractal theory space, which this paper applies to the Standard Model flavour sector.","marker":"[22]"},{"why":"Supplies the deconstruction method for writing mass-chain Lagrangians on a graph, used to build the fractal mass chain.","marker":"[23]"},{"why":"Provides the clockwork mechanism and the factorised $q_{ij}=f^{i-j}$ couplings that produce localised zero modes and exponential hierarchies.","marker":"[26]"},{"why":"Gives the node and edge labelling recurrences used to write the Hamiltonian on Sierpinski-like graphs at each iteration.","marker":"[30]"},{"why":"Supplies the pseudoinverse method used to integrate out the heavy modes and derive the effective $3\\times3$ Yukawa matrices.","marker":"[32–34]"},{"why":"Provides the experimental neutrino oscillation fit used to define the $3\\sigma$ target regions for mixing angles and mass orderings.","marker":"[36]"},{"why":"Provides the measured charged-lepton masses used as the targets for the mass-hierarchy plots.","marker":"[37]"}],"fun_headline_variants":["Fractal geometry yields three fermion generations","Three generations from a fractal mass chain","Fractal chain predicts lepton masses and mixing","Geometry decides how many generations exist","Fractal mass chains fix three Standard Model families"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the underlying network of Standard Model fermion mass terms is a Sierpinski-triangle fractal with exactly three decorated iterations, with couplings locked to the factorised clockwork pattern $q_{ij}=f^{i-j}$; neither the geometry nor the coupling pattern is derived from a deeper principle, and the identification of the three mathematical zero modes with the three observed generations is assumed.","fun_headline_variants_meta":{"raw":{"variants":["Fractal geometry yields three fermion generations","Three generations from a fractal mass chain","Fractal chain predicts lepton masses and mixing","Geometry decides how many generations exist","Fractal mass chains fix three Standard Model families"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000148,"raw_usage":{"total_tokens":1118,"prompt_tokens":806,"completion_tokens":312,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":422,"completion_tokens_details":{"reasoning_tokens":244}},"tokens_in":422,"tokens_out":312,"duration_ms":3649,"temperature":1.0,"reasoning_tokens":244,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-15T16:26:45.263750+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Run a global numerical scan over $f_L$, $f_E$, $f_N$ and the six Yukawa couplings in the ranges used in the paper, using the pseudoinverse formula (10) and the lepton mixing matrix formula (12), and demand that the resulting effective lepton Yukawa matrices simultaneously reproduce the three charged-lepton mass ratios, the two neutrino mass-squared differences for either ordering, and all three mixing angles inside their reported $3\\sigma$ ranges; if no parameter point satisfies all of these at once, the paper's claim that the framework reproduces the measured lepton masses and mixing angles is refuted.","supporting_citations":[{"cited_title":"Emergence of fractal geometries in the evolution of a metabolic enzyme.Nature, pages 1–7, 2024","cited_arxiv_id":null,"evidence_quote":"Introduced the concept of fractal theory space, which this paper applies to the Standard Model flavour sector."},{"cited_title":"Hofstadter Butterfly in Graphene","cited_arxiv_id":"2203.05821","evidence_quote":"Supplies the deconstruction method for writing mass-chain Lagrangians on a graph, used to build the fractal mass chain."},{"cited_title":"Fractal theory space: Spacetime of noninteger dimensionality.Physical Review D, 67(8):085004, 2003","cited_arxiv_id":null,"evidence_quote":"Provides the clockwork mechanism and the factorised $q_{ij}=f^{i-j}$ couplings that produce localised zero modes and exponential hierarchies."},{"cited_title":"A clockwork theory.Journal of High Energy Physics, 2017(2):1–39, 2017","cited_arxiv_id":null,"evidence_quote":"Gives the node and edge labelling recurrences used to write the Hamiltonian on Sierpinski-like graphs at each iteration."},{"cited_title":"The fourteenth western meeting of the american mathematical society, 1920","cited_arxiv_id":null,"evidence_quote":"Provides the experimental neutrino oscillation fit used to define the $3\\sigma$ target regions for mixing angles and mass orderings."},{"cited_title":"Application of calculus of matrices to method of least squares: with special reference to geodetic calculations.(No Title), 1951","cited_arxiv_id":null,"evidence_quote":"Provides the measured charged-lepton masses used as the targets for the mass-hierarchy plots."}],"review_version":2}