{"id":"b5634554-6f79-40da-9626-01c2ff9b07e0","arxiv_id":"2411.15663","paper_version":1,"verdict":"CONDITIONAL","confidence":"HIGH","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"For primes p not dividing n, the p-primary Tate-Shafarevich group over a dihedral extension of degree 2n has order equal, modulo fourth powers, to that over the quadratic subextension, or over the product of the three quadratic subextensions when n is even.","lead":"This note proves a formula for the size of the Tate-Shafarevich group of an elliptic curve over a dihedral extension of number fields, up to fourth powers and away from primes dividing the extension degree. The result constrains how the mysterious group Sha can grow when the base field is enlarged, which matters for understanding elliptic curves and their ranks.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The proof of Theorem 1 needs the p-primary Cassels–Tate pairing to justify squareness of the auxiliary p-primary Sha orders; the printed even-n congruence also has a subgroup typo.","rationale":"The central claim is supported by a correct combination of Brauer-relation isogenies and squareness of Tate–Shafarevich orders. The only substantive gap is that the squareness is applied to p-primary components using the full-group Cassels theorem, even though the theorem's hypothesis is finiteness of Sha(E/F)[p∞]. This is load-bearing because the proof's cancellation of |X(E/k)[p∞]|^2 and |X(E/F^{C2})[p∞]|^2 modulo fourth powers requires those p-primary orders to be squares; if one only knows the full-group theorem, the finiteness of the full groups is not available. However, the missing p-primary Cassels–Tate pairing statement is a standard and fixable omission, not a counterexample to the formula. The even-n displayed congruence has an evident typo (D_{2n} instead of D_n and D'_n), but the surrounding text and the Brauer relation in Lemma 6 make the intended argument clear, and the corrected congruence yields exactly the theorem's product formula. Because these are presentation issues rather than fatal mathematical errors, the reader's CONDITIONAL verdict is appropriate and should be retained; the proof should be revised to cite the p-primary Cassels–Tate pairing theorem and correct the subgroup notation in the even-n step.","tokens_in":4533,"tokens_out":44090,"duration_ms":397064,"concrete_test":"Verify analytically that for an elliptic curve E/L and a prime p, the Cassels–Tate pairing induces a nondegenerate alternating pairing on X(E/L)[p∞] whenever X(E/L)[p∞] is finite, and hence that |X(E/L)[p∞]| is a square. If this p-primary version is confirmed (it is standard), add a precise citation and recheck the squareness steps in the proof of Theorem 1. Then recompute the even-n congruence with F^{D_n} and F^{D'_n} in place of the printed F^{D_{2n}} and F^{D_{2n}'} to confirm the product formula for the three quadratic subextensions follows.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The main unstated assumption is in the squareness step of the proof of Theorem 1. Lemma 8 gives an equality of p-primary Tate–Shafarevich orders, e.g. for odd n, |X(E/F)[p∞]| |X(E/k)[p∞]|^2 = |X(E/K)[p∞]| |X(E/F^{C2})[p∞]|^2. The proof then says that because 'X of an elliptic curve is finite, it has square order', the extra factors can be discarded modulo fourth powers. But the theorem only assumes Sha(E/F)[p∞] is finite; the full groups over the subfields k, K, F^{C2}, and the quadratic subfields in the even case need not be finite. To conclude that |X(E/L)[p∞]| is a square for these subfields, one must invoke the p-primary Cassels–Tate pairing: when the p-primary component is finite, the Cassels–Tate pairing restricts to a nondegenerate alternating pairing on it, so its order is a square. The paper cites Cassels' theorem for finite Sha, but does not state or cite this p-primary version. Without it, the squareness steps in both the odd and even cases are unjustified as written. A secondary, more cosmetic issue is that the displayed congruence in the even-n case appears to write F^{D_{2n}} and F^{D_{2n}'} where the intended fields are F^{D_n} and F^{D'_n}, the fixed fields of the two order-n subgroups from Lemma 6; with the corrected subgroups the final formula does follow.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper proves a congruence modulo fourth powers for the size of the p-primary Tate-Shafarevich group of an elliptic curve in a dihedral extension F/k of degree 2n, for primes p not dividing n. Theorem 1 asserts that, assuming X(E/F)[p^\\infty] is finite, |X(E/F)[p^\\infty]|/|X(E/K)[p^\\infty]| is a fourth power when n is odd, and for even n the analogous ratio with the product of the three quadratic subextensions is a fourth power. The proof uses Brauer relations in the dihedral group to construct isogenies between products of Weil restrictions whose degrees are coprime to p, then applies a lemma equating p-primary Tate-Shafarevich orders and discards square factors using Cassels' squareness theorem. Theorem 4 gives a Z_p[H]-module decomposition X(E/F)[p^\\infty] \\cong X\\oplus X and X(E/K)[p^\\infty] \\cong X^H\\oplus X^H, with a squareness criterion for |X|/|X^H| under a congruence condition modulo n. The paper also gives Magma examples showing that the hypotheses on dihedrality and on p not dividing n are necessary.","tokens_in":99,"tokens_out":9149,"duration_ms":377100,"significance":"If Theorem 1 is established, it is a clean and useful structural result: it severely constrains the growth of the p-primary Tate-Shafarevich group in dihedral extensions for primes not dividing n, expressing it up to fourth powers in terms of the quadratic subextensions. The proof is mostly a transparent application of existing machinery (Bartel, Dokchitser-Dokchitser, Chetty, Cassels) rather than an introduction of fundamentally new ideas, and the explicit counterexamples for cyclic non-dihedral extensions and for p dividing n are valuable because they show the statement is sharp. The paper is honest about its limitations and cites its dependencies clearly. The main gap, the need for the p-primary Cassels-Tate squareness statement, is local and fixable; it does not affect the conceptual contribution.","major_comments":[{"comment":"The squareness step in the proof of Theorem 1 invokes 'When X of an elliptic curve is finite, it has square order' to discard factors such as |X(E/k)[p^\\infty]|^2 and |X(E/F^{C_2})[p^\\infty]|^2 in the odd case, and the analogous factors in the even case. But Theorem 1 assumes only that X(E/F)[p^\\infty] is finite; the footnote citing [5, Remark 2.10] gives finiteness of the p-primary groups over subfields, not squareness of their orders. The proof therefore needs the p-primary version of Cassels' theorem: if X(E/L)[p^\\infty] is finite, then the Cassels-Tate pairing restricts to a nondegenerate alternating pairing on it, so its order is a square. This version is not stated or cited in the paper. Without it, the step 'so |X(E/F)[p^\\infty]| \\equiv |X(E/K)[p^\\infty]| (mod Q^{*4})' is unjustified as written. The fix is local: add the p-primary statement as a lemma or a citation and apply it to each subfield.","section":"Proof of Theorem 1, Section 2"}],"minor_comments":[{"comment":"The displayed congruence in the even-n case uses F^{D_{2n}} and F^{D_{2n}'}; since D_{2n} is the whole group, F^{D_{2n}} = k, and D_{2n}' is undefined. The intended subgroups are D_n and D_n' from Lemma 6, so the congruence should read |X(E/F)[p^\\infty]| |X(E/F^{D_n})[p^\\infty]| \\equiv |X(E/F^{D_n'})[p^\\infty]| |X(E/F^{C_n})[p^\\infty]| (mod Q^{*4}). This typo should be corrected, as the current display does not follow from Lemma 8.","section":"Proof of Theorem 1, even-n case"},{"comment":"In the sentence 'This condition implies ord_p(d) is even for all d|n', the notation should be ord_d(p), the multiplicative order of p modulo d, rather than ord_p(d).","section":"Proof of Theorem 4, Section 3"},{"comment":"The two Magma examples are helpful, but the text should state explicitly that the computed quantities are analytic orders of Sha, so that the conclusion 'changes by a non-fourth-power' is understood to be conditional on the relevant cases of the Birch-Swinnerton-Dyer conjecture.","section":"Remark 2 and Remark 3"}],"recommendation":"major_revision","confidential_remarks":"The manuscript is a solid short note whose main theorem is likely correct, but the proof as written has one unstated assumption that must be fixed before publication: the p-primary Cassels-Tate squareness statement. The subgroup typo in the even-n proof is cosmetic but should be corrected in the same revision. Once these points are addressed, I would support acceptance."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Worth a look if you care about Sha growth in dihedral extensions. The note proves a modulo-fourth-power formula for |Sha(E/F)[p∞]| in terms of smaller fields when F/k is dihedral of degree 2n and p∤n: for odd n it matches the quadratic subfield, and for even n it matches the product of the three quadratic subfields. That is a genuinely new packaged statement, though every ingredient is known: Dokchitser–Dokchitser isogenies from Brauer relations, Bartel's Z_(p)-relation result, and Cassels squareness. The two numerical examples in Remarks 2 and 3 are a nice touch and show the hypotheses are not decorative.\n\nWhat the paper does well: Lemma 6 writes down the explicit dihedral Brauer relations, Lemmas 7 and 8 convert them into p-primary Sha order equalities via isogenies, and the proof of Theorem 1 is structurally sound. Theorem 4 is a reasonable bonus: under the same finiteness, it gives a Z_p[H]-module decomposition X(E/F)[p∞] ≅ X ⊕ X and identifies the K-invariants, with a square-ratio statement when p^a ≡ −1 mod n. The appeal to Chetty's lemma is terse but plausible.\n\nThe soft spots are real but minor. The proof of Theorem 1 says 'when X is finite it has square order' and applies that to the p-primary groups of the subfields k, F^{C2}, and (in the even case) the quadratic subfields. The theorem only assumes Sha(E/F)[p∞] finite; finiteness over the subfields is footnoted to Dokchitser–Dokchitser, but the square-order conclusion for the individual p-primary groups needs the p-primary Cassels–Tate pairing, which is not stated or cited. This is an easy fix—cite the p-primary version or note it explicitly—but as written it is an unstated assumption. Also, the displayed congruence in the even-n case has a typo: it writes F^{D_{2n}} and F^{D_{2n}'} where the intended fields are the fixed fields of the two order-n subgroups from Lemma 6, i.e. two of the three quadratic subfields. With the corrected subgroups the final formula does follow.\n\nBottom line: the central claim is correct in spirit and the gaps are fixable. This is a solid short note, not a landmark. I'd send it to a competent referee rather than desk-reject; the referee should ask for the p-primary Cassels statement and a corrected even-n congruence. I'd probably cite it if I were working on parity or arithmetic statistics in dihedral towers, and I might bring it to a reading group as a clean example of Brauer-relation isogenies in action.","headline":"A clean, largely correct short note that packages known Brauer-relation isogenies and Cassels squareness into a modulo-fourth-powers formula for Sha in dihedral extensions, with a couple of fixable rigour gaps.","tokens_in":5377,"tokens_out":7161,"would_cite":true,"duration_ms":58021,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["11G05","11G40"],"pacs":[],"model":"deepseek-v4-flash","headline":"For elliptic curves over dihedral extensions, the size of the Tate–Shafarevich group is fixed up to fourth powers by its sizes over quadratic subfields.","keywords":["Tate–Shafarevich group","elliptic curves","dihedral extensions","Brauer relations","Weil restriction","p-primary Sha","Galois module structure","square order"],"falsifier":"A direct descent computation of $\\mathrm{Sha}(E/F)[5^\\infty]$ for a curve over a dihedral extension of degree 6 (so $n=3$, $p=5 \\nmid n$) would settle Theorem 1: the ratio to the quadratic subfield's Sha order must be exactly a fourth power of 5, and any other pure power would disprove it.","tokens_in":4315,"feed_emoji":"🔢","tokens_out":9678,"duration_ms":76282,"temperature":0.7,"pith_summary":"This paper proves that for an elliptic curve over a number field, the size of the p-primary Tate–Shafarevich group over a dihedral extension $F/k$ of degree $2n$ is controlled up to fourth powers by the size over the quadratic subextensions, for any prime $p$ not dividing $n$. When $n$ is odd, the order over $F$ differs from the order over the unique quadratic subfield by a fourth power; when $n$ is even, it differs from the product of the orders over the three quadratic subfields by a fourth power. The proof uses Brauer relations in the dihedral group to build isogenies between products of Weil restrictions, transferring the classical square-order theorem for finite Tate–Shafarevich groups into a fourth-power comparison. A second theorem describes the Galois module structure of the p-primary Sha, showing it is a direct sum of two isomorphic submodules. This gives a new structural constraint on how Sha can grow in dihedral towers and a practical tool for computing it up to fourth powers.","feed_headline":"Dihedral extensions fix Sha growth up to fourth powers","feed_subtitle":"For primes away from the extension degree, the p-part of Sha over the big field is determined by quadratic subfields.","key_machinery":"The central object is a Brauer relation in the dihedral group $D_{2n}$: a formal sum of subgroups $\\sum H_i - \\sum H'_j$ whose permutation modules are isomorphic over $\\mathbb{Q}$. The paper uses two such relations, one for odd $n$ and one for even $n$, which become $\\mathbb{Z}_{(p)}$-relations for $p \\nmid n$. These induce isogenies between products of Weil restrictions of $E$ to the fixed fields, of degree coprime to $p$, so the p-primary Tate–Shafarevich orders of the two products are equal. Comparing the factors and applying the square-order theorem for finite Sha gives the fourth-power formula. For the module structure, the key is that $\\mathbb{Z}_p[C_n]$ splits as a product of local rings with principal maximal ideals, which forces the p-primary Sha over $F$ to be a direct sum of two isomorphic modules.","core_discovery":"Let $E$ be an elliptic curve over a number field $k$ and let $F/k$ be a dihedral extension of degree $2n$. Fix a prime $p$ not dividing $n$ and assume the p-primary Tate–Shafarevich group $\\mathrm{Sha}(E/F)[p^\\infty]$ is finite. The paper establishes that for odd $n$, $|\\mathrm{Sha}(E/F)[p^\\infty]| / |\\mathrm{Sha}(E/K)[p^\\infty]| = p^{4t}$ for some integer $t$, where $K$ is the quadratic subextension; for even $n$, $|\\mathrm{Sha}(E/F)[p^\\infty]|$ equals the product $|\\mathrm{Sha}(E/K_1)[p^\\infty]| |\\mathrm{Sha}(E/K_2)[p^\\infty]| |\\mathrm{Sha}(E/K_3)[p^\\infty]|$ times $p^{4t}$, where $K_1, K_2, K_3$ are the three quadratic subextensions. The argument works by applying dihedral Brauer relations to obtain isogenies of Weil restrictions with degree coprime to $p$, so the two sides have identical p-primary Sha orders, and then using the fact that a finite Sha has square order to compare the extra factors. The same relations yield a Galois module statement: over the cyclic subgroup $H$ of order $n$, $\\mathrm{Sha}(E/F)[p^\\infty] \\cong X \\oplus X$ and $\\mathrm{Sha}(E/K)[p^\\infty] \\cong X^H \\oplus X^H$, and in the odd case with $p^a \\equiv -1 \\bmod n$ solvable, $|X|/|X^H|$ is a square.","pith_inferences":["A likely next step is to supply the missing p-primary Cassels squareness theorem explicitly; once stated, the proof of Theorem 1 becomes self-contained and could be adapted to other p-primary pairings.","The same Brauer-relation technique might apply to other non-dihedral groups whose rational permutation modules coincide, producing analogous fourth-power formulas for other solvable extensions.","One testable prediction: for a fixed curve and a dihedral tower with $p$ not dividing any of the degrees, the $p$-adic valuation of $|\\mathrm{Sha}|$ modulo 4 should follow a deterministic pattern determined by the quadratic subfields; a computational search over quadratic twists could confirm or refute this across many examples.","If the analytic order of Sha is used via the Birch–Swinnerton-Dyer conjecture, the fourth-power formula translates into a constraint on the special values of $L$-functions of elliptic curves over dihedral extensions."],"forward_implications":["For odd dihedral extensions, the p-primary part of Sha over the full field is, up to a fourth power, the same as over the quadratic subfield; so any new Sha in the extension is constrained to a very specific size.","For even dihedral extensions, the same holds with the three quadratic subfields playing the role of the base; the product of their Sha orders determines the big Sha up to fourth powers.","The Galois module splitting $\\mathrm{Sha}(E/F)[p^\\infty] \\cong X \\oplus X$ means the group is even as a $\\mathbb{Z}_p[H]$-module, and the fixed part is the same module twice, giving constraints on possible module structures beyond mere sizes.","When $n$ is odd and $p^a \\equiv -1 \\pmod{n}$ has a solution, the sharper ratio $|X|/|X^H|$ is a square, so the fourth-power formula can be upgraded to a square condition.","The result is sharp: the paper gives explicit examples showing the formula fails for general cyclic extensions of odd degree and when $p$ divides $n$."],"supporting_citations":[{"why":"Supplies the key theorem that Brauer relations induce isogenies between products of Weil restrictions, the foundation of the proof.","marker":"[5]"},{"why":"Gives the $\\mathbb{Z}_{(p)}$-relation and lattice-inclusion argument showing the induced isogeny has degree coprime to $p$.","marker":"[1]"},{"why":"States the resulting isogeny degree $d^2$, transferring the comparison from isogenies to Tate–Shafarevich orders.","marker":"[2]"},{"why":"Provides the square-order theorem for finite Tate–Shafarevich groups used to reduce the comparison to fourth powers.","marker":"[6]"},{"why":"Used in the chain (via [2]) for the isogeny degree and its effect on Tate–Shafarevich groups of abelian varieties.","marker":"[7]"},{"why":"Supplies the pairing lemma used to prove the Galois module splitting $\\mathrm{Sha}(E/F)[p^\\infty] \\cong X \\oplus X$.","marker":"[3]"}],"fun_headline_variants":["Dihedral Sha: quadratic subfields determine up to fourth powers","Sha in dihedral extensions: fourth power precision from quadratic subfields","Dihedral extensions: Sha order from quadratic subfields, up to quartic","Quadratic subfields control Sha in dihedral extensions to fourth powers"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The argument depends on the p-primary square-order theorem for finite Tate–Shafarevich groups, but the paper assumes finiteness only for $\\mathrm{Sha}(E/F)[p^\\infty]$ and does not state or cite the p-primary version it applies.","fun_headline_variants_meta":{"raw":{"variants":["Dihedral Sha: quadratic subfields determine up to fourth powers","Sha in dihedral extensions: fourth power precision from quadratic subfields","Dihedral extensions: Sha order from quadratic subfields, up to quartic","Quadratic subfields control Sha in dihedral extensions to fourth powers"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000954,"raw_usage":{"total_tokens":4085,"prompt_tokens":980,"completion_tokens":3105,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":596,"completion_tokens_details":{"reasoning_tokens":3029}},"tokens_in":596,"tokens_out":3105,"duration_ms":21299,"temperature":1.0,"reasoning_tokens":3029,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T14:05:42.706124+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A direct descent computation of $\\mathrm{Sha}(E/F)[5^\\infty]$ for a curve over a dihedral extension of degree 6 (so $n=3$, $p=5 \\nmid n$) would settle Theorem 1: the ratio to the quadratic subfield's Sha order must be exactly a fourth power of 5, and any other pure power would disprove it.","supporting_citations":[{"cited_title":"Dokchitser, V.Dokchitser On the Birch–Swinnerton-Dyer quotients modulo squares","cited_arxiv_id":null,"evidence_quote":"Supplies the key theorem that Brauer relations induce isogenies between products of Weil restrictions, the foundation of the proof."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Gives the $\\mathbb{Z}_{(p)}$-relation and lattice-inclusion argument showing the induced isogeny has degree coprime to $p$."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"States the resulting isogeny degree $d^2$, transferring the comparison from isogenies to Tate–Shafarevich orders."},{"cited_title":"Cassels Arithmetic on Curves of Genus 1","cited_arxiv_id":null,"evidence_quote":"Provides the square-order theorem for finite Tate–Shafarevich groups used to reduce the comparison to fourth powers."},{"cited_title":"Milne On the Arithmetic of Abelian Varieties","cited_arxiv_id":null,"evidence_quote":"Used in the chain (via [2]) for the isogeny degree and its effect on Tate–Shafarevich groups of abelian varieties."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the pairing lemma used to prove the Galois module splitting $\\mathrm{Sha}(E/F)[p^\\infty] \\cong X \\oplus X$."}],"review_version":1}