{"id":"efa69c7b-3864-4c8c-a698-0e74c9dea7dc","arxiv_id":"2412.20606","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"The thesis extends algebraic 'newform' definitions to characteristic p modular forms with character, using Hecke kernels and the Fricke operator.","lead":"This mathematics thesis studies algebraic ways to define 'newforms' for modular forms in characteristic p, building on work by Deo and Medvedovsky. It extends their definitions to modular forms with character and to non-squarefree levels, with several new theorems and explicit examples.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The reader's cited Fricke-integrality concern targets Thm 6.21, not the stated central claim Thm 6.18; the load-bearing gap for Thm 6.18 is the unproved character-twisted oldform decomposition used in its proof sketch.","rationale":"I read Theorem 6.18 as the central claim. I checked the algebra of the character-twisted operators: with ell T_ell = ell U_ell + chi(ell)W_ell, U_ell W_ell = S_ell and chi(ell)=chi'(ell), the displayed characterization in the proof of Theorem 6.18 is formally consistent, and the characteristic polynomial in Theorem 6.13 also matches. The genuine soft spot is that the proof of Theorem 6.18 is a sketch: it invokes a character-twisted analogue of Proposition 5.9, which in turn requires the ell-old cusp space with character to equal S^{chi'}(N/ell,B) ⊕ W_ell(S^{chi'}(N/ell,B)) over characteristic-p coefficient domains. That equality is proved only in characteristic zero (Proposition 4.5) and for trivial character in Proposition 5.5; no proof is supplied for nontrivial chi and non-flat B. The reader's cited Fricke integrality is a real issue for Theorem 6.21, but it is not load-bearing for Theorem 6.18. A dimension check over F_p in a small level and a full re-derivation of the twisted Proposition 5.9 would settle whether the gap is substantive. This does not lower my confidence below the reader's conditional acceptance; it sharpens where the verification should focus.","tokens_in":44335,"tokens_out":30601,"duration_ms":284082,"concrete_test":"Independently prove the character-twisted analogue of Proposition 5.9, including the direct-sum identity, for B = F_p[chi]. As a numerical check, take p=5, N=10, ell=2, chi a nontrivial character of conductor 5, weight k=4, and compute in SageMath the dimensions of S^chi_4(10,F_5)_{2-old} from the tensor definition and of S^{chi'}_4(5,F_5) ⊕ W_2(S^{chi'}_4(5,F_5)). If the dimensions disagree, the decomposition used in the proof of Theorem 6.18 fails; if they agree, repeat for p=7, N=21, ell=3, chi conductor 7, k=4. A full re-derivation should also verify that the two conditions ell T_ell f = -(ell+1)S_ell g and ell T_ell g = -(ell+1)chi'(ell) f are stable under reduction modulo p^{b+1} in the lifting step.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The stated central claim, Theorem 6.18, does not use the scaled Fricke operator H_N; the integrality cited to [Loe] enters only Definition 6.20 and Theorem 6.21. The load-bearing step for Theorem 6.18 is the 'slight modification of Proposition 5.9' invoked in its proof. That modification characterizes elements f + W_ell(g) of the ell-old subspace, so it presupposes that, for B a Z[1/N,chi]-domain of characteristic p, S^chi_k(N,B)_{ell-old} = S^{chi'}_k(N/ell,B) ⊕ W_ell(S^{chi'}_k(N/ell,B)). Section 6.1 defines S^chi_k(N,B)_{ell-old} by extending scalars from Z[chi]-integral oldforms; the only proof of the direct-sum description is Proposition 4.5, whose final tensor step uses flatness over Z[1/ell] and hence does not apply to characteristic-p B. Proposition 5.5 gives the analogous statement only for the trivial character. If this direct-sum decomposition fails for some nontrivial chi, the modified Proposition 5.9 and the p-adic lifting argument do not apply to the tensor-defined old subspace, and Theorem 6.18 is not established.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The manuscript is an MSc thesis that develops algebraic definitions of newforms in characteristic p, following Deo and Medvedovsky. Chapters 1–3 review classical Atkin–Lehner theory and the squarefree-level Uℓ-new and Trℓ-new notions. Chapter 4 sets up modular forms over commutative rings and proves that in characteristic zero the algebraic notions coincide with classical newforms. Chapter 5 treats characteristic p and proves (Theorem 5.10) that for a Z[1/ℓ]-domain B of characteristic p, S(N,B)ℓ-new ⊆ S(N,B)Uℓ-new = S(N,B)Trℓ-new. Chapter 6 generalizes to modular forms with Dirichlet character, defining Uχℓ-new and Trχℓ-new and claiming equality in characteristic p (Theorem 6.18); it also proposes a Fricke-operator variant for non-squarefree levels (Theorem 6.21). The new part of the thesis is Chapter 6, where several proofs are presented only as sketches.","tokens_in":44629,"tokens_out":11184,"duration_ms":106102,"significance":"If the claims are correct, Theorem 6.18 is a natural extension of Deo–Medvedovsky's theorem to forms with character, and the framework provides useful algebraic replacements for classical newforms in characteristic p. The thesis is valuable as an exposition: it fills in many details of [DM19], gives explicit computations and worked examples, and clearly identifies the places where classical theory breaks down. However, the genuinely new part of Chapter 6 rests on several proof sketches and on an externally cited integrality statement, so the significance is conditional on completing those arguments.","major_comments":[{"comment":"The proof of Theorem 6.18 is not complete as written. It invokes 'a slight modification of Proposition 5.9' for character-twisted spaces, but that proposition presupposes a direct-sum description Sχ_k(N,B)ℓ-old = Sχ′_k(N/ℓ,B) ⊕ Wℓ(Sχ′_k(N/ℓ,B)), or at least trivial intersection of the two summands. Section 6.1 defines Sχ_k(N,B)ℓ-old by base change from Z[χ]-integral oldforms, and the only direct-sum statements proved earlier (Proposition 4.5 and Proposition 5.5(b)) are for the trivial character. No character-twisted analogue of Lemma 4.6/Proposition 4.5 is stated or proved. If the intersection is nontrivial for some nontrivial χ, the displayed description of (Sχ_k(N,B)ℓ-old)Uχℓ−new is unjustified and the p-adic lifting argument in Theorem 6.18 has no foundation. This is load-bearing because Theorem 6.18 is the central new result.","section":"§6.4, Theorem 6.18"},{"comment":"The characteristic-zero identifications Sχ(N,B)ℓ-new = Sχ(N,B)Uχℓ−new and Sχ(N,B)ℓ-new = Sχ(N,B)Trχℓ−new are asserted by saying that one repeats the proofs of Theorems 3.9/3.14 'whilst keeping track of character'. In Theorem 6.13 the reduction to B=C is justified only by flatness of B over Z[χ]; flatness alone does not transfer an equality of kernels unless one has already established the relevant base-change compatibility and Z[χ]-integral bases for the spaces and kernels. Proposition 6.2 provides bases, but the kernel/base-change step is not written. Since Theorem 6.18's p-adic argument uses the characteristic-zero equality over Q_p[χ] via its lifting step, this omitted verification is also load-bearing for the central claim.","section":"§6.3–6.4, Theorems 6.13 and 6.17"},{"comment":"Property (iv) of the Fricke operator — that h_N and H_N extend to Mχ_k(N,B) for any Z[1/N,ζ_d,χ]-algebra B — is cited only to a MathOverflow answer [Loe] and is not proved in the thesis. This is not a purely cosmetic gap: Definition 6.20 and Theorem 6.21 are formulated in terms of Trχℓ and H_N, and the equality Sχ(N,B)ℓ-new = Sχ(N,B)Trχℓ−new′ depends on the extension of H_N to the coefficient ring B. The footnote in §6.2 even records that the author is unaware of a proof of the expected stronger Z[1/ℓ,ζ_m]-integrality. For a journal submission, the integrality statement cited to [Loe] needs to be either proved or replaced by a complete peer-reviewed reference. I note that this gap does not bear on Theorem 6.18, since that theorem avoids H_N.","section":"§6.5, Definition 6.20 and Theorem 6.21"}],"minor_comments":[{"comment":"The displayed formula 'Tℓ = Uℓ + ℓSℓWℓf' appears to be a typo; the subsequent argument in Proposition 5.9 uses the correct relation ℓUℓ = ℓTℓ − Wℓ, consistent with Tℓ = Uℓ + ℓ^{-1}Wℓ. Please correct the formula.","section":"§5.2"},{"comment":"The acknowledgement that the expected Z[1/ℓ,ζ_m]-integrality of the Atkin–Lehner operator is unproved should be moved into the main text, since it directly affects the scope and reliability of the non-squarefree generalization in Section 6.5.","section":"§6.2, footnote (i)"},{"comment":"The notation S(N,B) is used both for the sum over weights and for the algebra of cusp forms; in characteristic p this sum is not direct, so statements such as S(N,B)ℓ-new ⊆ S(N,B)Uℓ−new in Theorem 5.10 should explicitly state whether they are made in each graded piece or in the non-direct sum. The proof argues in fixed weight, but the theorem statements would benefit from an explicit convention.","section":"Notation throughout"}],"recommendation":"major_revision","confidential_remarks":"The manuscript is clearly an MSc thesis and reads as such. The exposition of the known material is solid, but the original Chapter 6 is at the level of proof sketches. For a journal, I would want the character-twisted oldform decomposition and the Fricke integrality either proved or cited to a peer-reviewed source. The MathOverflow citation [Loe] is not acceptable as the sole support for a load-bearing statement. I would also ask the authors to compare the claimed generalization with the existing literature on newforms with character in characteristic p, since the thesis only cites [DM19] for the trivial-character case."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Chapters 3–5 are a dependable, fully written-out account of Deo and Medvedovsky’s algebraic newforms in squarefree level and characteristic p. The exposition is careful, and it catches a real oversight in DM19: in weight 2 the Eisenstein space can make Uℓ-new and Trℓ-new disagree with the classical ℓ-new space (Example 3.18). That part is solid and useful on its own.\n\nChapter 6 is the original contribution: modular forms with character, the operators D^χ_ℓ and Tr^χ_ℓ, the equality S^χ(N,B)^{U^χ_ℓ-new} = S^χ(N,B)^{Tr^χ_ℓ-new} in characteristic p (Theorem 6.18), and a Fricke-operator version aimed at non-squarefree levels (Section 6.5). The author says at the start of Chapter 6 that most proofs will be short; that warning is accurate, and it marks the vulnerable spot.\n\nThe stress-test note is right to separate Theorem 6.18 from the Fricke integrality. Theorem 6.18 does not use H_N. It needs a “slight modification of Proposition 5.9” for the character-twisted ℓ-old subspace. That proposition describes elements f + W_ℓ(g) as sitting in the old space, so it presupposes S^χ_k(N,B)_ℓ-old = S^{χ′}_k(N/ℓ,B) ⊕ W_ℓ(S^{χ′}_k(N/ℓ,B)). For nontrivial χ, Section 6.1 defines the old space by extension of scalars from Z[χ]-integral oldforms. The direct-sum description is proved only in the trivial-character case (Proposition 5.5) using flatness over Z[1/ℓ], and that argument does not carry to characteristic-p coefficient rings. If the decomposition fails for some nontrivial character, the modified Proposition 5.9 and the p-adic lifting argument do not apply to the tensor-defined old subspace, and Theorem 6.18 is not established. This is a real gap in the central new claim, not a stylistic quibble.\n\nThe other issue is smaller: the integrality of the scaled Fricke operator in Section 6.5 is cited to a MathOverflow answer [Loe]. That underpins Definition 6.20 and Theorem 6.21, not 6.18, so it is not what threatens the main theorem, but it is still a shaky foundation for a claimed generalization.\n\nThe citation pattern is fair; DM19 and Li are used appropriately, and the thesis is explicit about which parts are expository and which parts are new. There is no machine-checked proof or reusable code here, so the claims stand on ordinary prose proofs, which is exactly why the sketched parts matter.\n\nThe audience is clear: graduate students and researchers using mod p Hecke algebras, Monsky-style filtrations, or newform congruences will get real value from Chapters 3–5 and from Chapter 6 as a roadmap. I would send it to a serious referee, with the expectation that the author fills in the character-twisted old-space proof and replaces the MathOverflow citation with a proof. Conditional acceptance, not desk reject.","headline":"A genuinely useful write-up of Deo–Medvedovsky plus a plausible but under-proved new chapter on character-twisted newforms; send it to a referee who will ask for the old-space gap to be filled.","tokens_in":45155,"tokens_out":4653,"would_cite":true,"duration_ms":46212,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["11F11","11F25","11F33"],"pacs":[],"model":"deepseek-v4-flash","headline":"Cusp forms with Dirichlet character over characteristic-p coefficient rings have the same ℓ-new subspaces whether newness is defined by the U_ℓ-eigenvalue operator or by the kernels of trace operators.","keywords":["newforms","characteristic p","modular forms with character","Atkin-Lehner operator","Fricke operator","U-ell-new","trace operator","Dirichlet character"],"falsifier":"To test the central equality directly, compute the dimensions of $\\ker(D^\\chi_\\ell)$ and $\\ker(\\mathrm{Tr}^\\chi_\\ell) \\cap \\ker(\\mathrm{Tr}^\\chi_\\ell W_\\ell)$ inside $S^\\chi_k(N,\\mathbb{F}_p)$ for a small case such as $N=22$, $\\ell=2$, $\\chi$ induced by the quadratic character mod $11$, and $p=3$, in a weight where the space is nonzero; if the two dimensions differ, Theorem 6.18 is false, and if they agree, the theorem survives this test.","tokens_in":44135,"feed_emoji":"🧮","tokens_out":11389,"duration_ms":107644,"temperature":0.7,"pith_summary":"This thesis addresses a basic problem: in characteristic p, the classical decomposition of cusp forms into old and new parts breaks because of congruences, and the inner-product definition of newform no longer behaves. The paper takes up an algebraic way of defining newforms, using kernels of operators built from Hecke and Atkin-Lehner (and later Fricke) operators, and asks whether the two main algebraic notions agree for forms with Dirichlet character. Its central result is that they do: whenever ℓ divides the level exactly once and the character is induced from level $N/\\ell$, the space $S^\\chi(N,B)^{U^\\chi_\\ell\\text{-new}}$ equals $S^\\chi(N,B)^{\\mathrm{Tr}^\\chi_\\ell\\text{-new}}$ for any coefficient domain $B$ of characteristic $p$ containing $1/N$ and the values of $\\chi$. In characteristic zero the same algebraic notions are shown to reproduce the classical newforms with character, and the trace definition is extended to levels where ℓ divides N more than once by replacing the Atkin-Lehner operator with the scaled Fricke operator. A sympathetic reader should care because this supplies a definition of newform that survives in characteristic p and is available for a wider class of levels.","feed_headline":"Two algebraic newform definitions coincide in characteristic p","feed_subtitle":"With Dirichlet character, U-ℓ and trace kernels give the same newforms; a Fricke-operator route covers non-squarefree levels.","key_machinery":"The load-bearing objects are the modified Hecke operator $D^\\chi_\\ell = \\ell^2 U_\\ell^2 - \\chi'(\\ell)\\ell^k$ and the trace operators $\\mathrm{Tr}^\\chi_\\ell(f) = f + \\ell^{1-k}U_\\ell W_\\ell f$ (when $\\ell$ divides $N$ exactly once) and $\\mathrm{Tr}^\\chi_\\ell(f) = \\chi(-1)\\ell N^{-k}H_{N/\\ell}U_\\ell H_N f$ (in general), where $W_\\ell = \\ell^{k/2}w_\\ell$ is the scaled Atkin-Lehner operator and $H_N = N^{k/2}h_N$ is the scaled Fricke operator. The character enters through the twisted eigenvalue condition $U_\\ell^2 = \\chi'(\\ell)\\ell^{k-2}$ and through the operator identity $\\ell T_\\ell = \\ell U_\\ell + \\chi(\\ell)W_\\ell$ on forms coming from level $N/\\ell$; these identities are what make the kernels of $D^\\chi_\\ell$, $\\mathrm{Tr}^\\chi_\\ell$, and $\\mathrm{Tr}^\\chi_\\ell W_\\ell$ comparable. The p-adic lifting argument is the engine that transfers kernel equalities from characteristic zero down to characteristic $p$.","core_discovery":"On the paper's own terms, the central claim is that 'new' can be defined purely algebraically in characteristic p, and that the two algebraic definitions coincide for cusp forms with character. The U definition declares $U^\\chi_\\ell$-new to be the kernel of $D^\\chi_\\ell = \\ell^2 U_\\ell^2 - \\chi'(\\ell)\\ell^k$, so a form is new when $U_\\ell^2 f = \\chi'(\\ell)\\ell^{k-2}f$; the trace definition declares $\\mathrm{Tr}^\\chi_\\ell$-new to be the simultaneous kernel of $\\mathrm{Tr}^\\chi_\\ell$ and $\\mathrm{Tr}^\\chi_\\ell W_\\ell$. Theorem 6.18 proves these agree on $S^\\chi(N,B)$ whenever $\\chi$ is induced by a character mod $N/\\ell$ and $B$ is a $Z[1/N,\\chi]$-domain of characteristic $p$. The proof reduces the field case to $B = \\mathbb{F}_p$, lifts a form in the $U$-kernel to $Z_p[\\chi]$ with a $p$-adic decomposition into old and new parts, clears denominators, and pushes the kernel condition down modulo powers of $p$. The paper also shows in characteristic zero that the algebraic definitions agree with the classical $\\ell$-new spaces, and it extends the trace definition to non-squarefree levels using the scaled Fricke operator, where the characteristic-zero agreement with classical newforms is proved via Li's theorem.","pith_inferences":["If the missing Fricke integrality is supplied, the same p-adic lifting proof should yield the non-squarefree characteristic-$p$ equality, making the definition of newform uniform across all levels.","The equality of kernels suggests a canonical metric-free 'new' submodule of the Hecke module $S^\\chi(N,B)$, which could serve as the correct input for studying Hecke-stable filtrations in characteristic $p$.","A computational scan of small $N$, $\\ell$, $p$, and characters could map where the inclusion $S^\\chi(N,B)^{\\ell\\text{-new}} \\subseteq S^\\chi(N,B)^{U^\\chi_\\ell\\text{-new}}$ is strict, linking strictness to the congruence constant $(\\ell+1)\\ell^{(k-2)/2}\\chi'(\\ell)$.","The reliance on an outside reference for Fricke integrality points to a concrete missing lemma: proving $H_N$ integrality on $q$-expansion bases over $Z[1/N,\\zeta_d,\\chi]$ would put the non-squarefree story on the same footing as the squarefree case."],"forward_implications":["In squarefree level, cusp forms with character have a Hecke-theoretic definition of $\\ell$-newform that is independent of the Petersson inner product, with the $U_\\ell$ and trace formulations coinciding in characteristic $p$.","The trace/Fricke formulation gives a candidate definition of $\\ell$-newform at non-squarefree levels; in characteristic zero it provably matches the classical $\\ell$-new spaces, so characteristic-$p$ computations have a known characteristic-zero benchmark.","The explicit description of old forms lying in the $U$/trace kernels, via the equations $\\ell T_\\ell f = -(\\ell+1)S_\\ell g$ and $\\ell T_\\ell g = -(\\ell+1)\\chi'(\\ell)f$, makes the failure of the naive old/new decomposition mod $p$ controllable.","Because the equality is proved by flatness and kernel preservation, it passes to every $Z[1/N,\\chi]$-domain of characteristic $p$, so the algebraic theory is base-change friendly.","In characteristic zero, the algebraic definitions with character agree with the classical $\\ell$-new spaces even when $\\ell^2$ divides $N$, extending Li's characterization to modules over $Z[1/N,\\zeta_d,\\chi]$-domains."],"supporting_citations":[{"why":"Supplies the two algebraic definitions of newness—$U_\\ell$-new and $\\mathrm{Tr}_\\ell$-new—and the squarefree-level characteristic-p equality that this thesis generalizes to forms with character.","marker":"[DM19]"},{"why":"Origin of the classical newform theory and of the Atkin-Lehner and Hecke operator identities that motivate the algebraic definitions.","marker":"[AL70]"},{"why":"Provides the character-twisted theory of newforms, including the eigenvalue condition $U_\\ell^2 = \\chi'(\\ell)\\ell^{k-2}$ and the trace/Fricke characterization used in the chapter on forms with character.","marker":"[Li75]"},{"why":"Supplies the Fricke-based trace formula for levels where $\\ell$ divides $N$ more than once, on which Definition 6.20 and Theorem 6.21 rest.","marker":"[Wei77]"},{"why":"Cited as the source for integrality of the Fricke/Atkin-Lehner operator over $Z[1/N,\\zeta_d,\\chi]$, the unproven input needed for the non-squarefree extension.","marker":"[Loe]"},{"why":"Provides the Weil bound used in the squarefree proofs to rule out oldforms with the forbidden $U_\\ell^2$ eigenvalue $\\ell^{k-2}$.","marker":"[Del74]"}],"fun_headline_variants":["Algebraic newform definitions match in char p for cusp forms","U-l and trace kernels coincide for char-p newforms","Char-p newforms: two algebraic definitions agree","Algebraic newform coincidence proven via p-adic lifting","Newform algebra in char p: U and trace kernels align"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The argument assumes, on the authority of an outside reference rather than a proof, that the scaled Fricke operator $H_N$ preserves coefficient rings of the form $Z[1/N,\\zeta_d,\\chi]$; if that integrality fails, the extension to non-squarefree levels and the equality in Theorem 6.21 lose their base ring.","fun_headline_variants_meta":{"raw":{"variants":["Algebraic newform definitions match in char p for cusp forms","U-l and trace kernels coincide for char-p newforms","Char-p newforms: two algebraic definitions agree","Algebraic newform coincidence proven via p-adic lifting","Newform algebra in char p: U and trace kernels align"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001154,"raw_usage":{"total_tokens":4772,"prompt_tokens":927,"completion_tokens":3845,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":543,"completion_tokens_details":{"reasoning_tokens":3773}},"tokens_in":543,"tokens_out":3845,"duration_ms":28594,"temperature":1.0,"reasoning_tokens":3773,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-10T23:15:33.448019+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"To test the central equality directly, compute the dimensions of $\\ker(D^\\chi_\\ell)$ and $\\ker(\\mathrm{Tr}^\\chi_\\ell) \\cap \\ker(\\mathrm{Tr}^\\chi_\\ell W_\\ell)$ inside $S^\\chi_k(N,\\mathbb{F}_p)$ for a small case such as $N=22$, $\\ell=2$, $\\chi$ induced by the quadratic character mod $11$, and $p=3$, in a weight where the space is nonzero; if the two dimensions differ, Theorem 6.18 is false, and if they agree, the theorem survives this test.","supporting_citations":[],"review_version":1}