{"id":"78a07ba6-fd88-4b66-80d1-e6c6af6137b1","arxiv_id":"1908.09197","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"The paper upgrades Howard's divisibility to the full Howard main conjecture when the Heegner-point Kolyvagin system is primitive, and shows the conjecture is equivalent to primitivity plus a Tamagawa-factor condition.","lead":"This paper proves Howard's main conjecture, a central prediction in the arithmetic of elliptic curves, in new cases including elliptic curves with analytic rank above one. The key is a refined counting tool, the Kolyvagin system error term, that turns a known divisibility into an equality.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The bridge congruence from κ primitive to κHg primitive in Propositions 3.2.2–3.2.3 is asserted tersely; a p-factor error there would break Theorem A.","rationale":"The reader's weakest-assumption analysis identifies the same point I consider most load-bearing: Propositions 3.2.2 and 3.2.3 are the unique bridge from the hypothesis 'κ is primitive' to the Λ-primitivity of κ^Hg, and hence to the equality in Howard's Main Conjecture. The surrounding formalism—d(κ), the error term in Theorem 2.3.6, and the deformation argument in Theorem 2.4.5—is internally coherent, and I found no circularity: Theorem B is explicitly conditional on Howard's Main Conjecture, and Theorem C is a genuine equivalence rather than a disguised assumption. There are no fitted parameters, and the paper gives substantial credit to the external results it invokes (Howard, Castella–Wan, Jetchev–Skinner–Wan, CH18, Wei Zhang). The main risk is therefore verification of the short local computations in Section 3.2. I also noticed an apparent sign discrepancy in the inert ordinary congruence of Proposition 3.2.2—direct reduction of the displayed Φ gives (1−a_p^2) rather than (a_p^2−1)—which is harmless for primitivity but reinforces that these identities deserve an independent check. My concrete test is designed to settle the bridge directly: recompute the scalar factors and verify that a nonzero finite-level class produces a nonzero class in the Iwasawa Selmer structure. Until that check is performed, CONDITIONAL is the appropriate verdict; no change to the reader's verdict is needed.","tokens_in":35556,"tokens_out":13759,"duration_ms":145487,"concrete_test":"Independently recompute the three congruences from the definitions in Section 3.2. For ordinary split p, expand Φ = (p−a_pσ+σ^2)(p−a_pσ*+σ*^2), use D0σ = D0 = D0σ* in the group ring of Gal(K[1]/K), and reduce modulo p; for ordinary inert p, reduce Φ = (p+1)^2 − a_p^2 and record whether the scalar is 1−a_p^2 or a_p^2−1; for supersingular split p, expand γ0 = a_p − σ − σ* and use a_p ≡ 0 mod p. Then, taking a core vertex m with κ_m ≠ 0 in E(K[m])/I_m modulo p, trace the Kummer image of D0DmΦP[m] (respectively D0Dmγ0P[m]) into H^1_{FΛ(m)}(K,T/m) and check both that it is nonzero and that it satisfies the Kolyvagin-system relation of Definition 2.1.1. If the scalar is a p-adic unit in all three cases and the nonzero finite-level class survives the Kummer and Selmer-condition maps, Propositions 3.2.2–3.2.3 are confirmed and Theorem A goes through.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The entire upgrade from Howard's divisibility to equality rests on the implication κ primitive ⇒ κ^Hg primitive, established only by the congruences in Propositions 3.2.2 and 3.2.3: in the ordinary split case D0DmΦP[m] ≡ (a_p−1)^2D0DmP[m] mod p, in the ordinary inert case D0DmΦP[m] ≡ (a_p^2−1)D0DmP[m] mod p, and in the supersingular split case D0Dmγ0P[m] ≡ −2D0DmP[m] mod p. These are the only place where the finite-level primitivity of κ is transported to the Iwasawa-level Kolyvagin system κ^Hg; if any displayed scalar were off by a p-factor, κ^Hg would not be Λ-primitive and Theorem 3.3.1 would give only Howard's old divisibility, not the equality in Theorem A. The derivations are compressed into single lines, no appendix revisits them, and the use of D0σ_p = D0σ*_p = D0 and the passage from a nonzero D0DmP[m] modulo p to a nonzero class in H^1_{FΛ(m)}(K,T/m) is not written out. Notably, in the inert ordinary case, directly reducing Φ = (p+1)^2 − a_p^2 modulo p gives (1−a_p^2)D0DmP[m], not (a_p^2−1)D0DmP[m]; the sign is a p-adic unit and thus harmless for primitivity, but it illustrates exactly the type of factor error that would be fatal. A missing or incorrect unit in any of the three congruences would sever the chain κ primitive ⇒ κ^Hg primitive ⇒ Λ-primitive ⇒ equality in Howard's Main Conjecture.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies Howard's main conjecture for Heegner points in the anticyclotomic Iwasawa theory of an elliptic curve. Its central result, Theorem A, says that if the finite-level Heegner Kolyvagin system κ is primitive and the usual running hypotheses hold (including the non-anomalous condition), then Howard's Main Conjecture is true, i.e. the divisibility obtained by Howard is upgraded to equality and the Selmer module has the predicted pseudo-isomorphism type. Theorem B gives a conditional converse: assuming Howard's Main Conjecture and a split-ordinary or analytic-rank-one situation, κ is primitive exactly when the split Tamagawa factors are p-indivisible. Theorem C combines these into an equivalence between primitivity of κ, Howard's Main Conjecture, and the absence of split Tamagawa factors. The main new technical ingredients are an improvement of Howard's Kolyvagin-system formalism with an error term d(κ) satisfying length(M)=length(H^1_F(K,T)/Rκ_1)-d(κ), and a comparison between the finite-level and Iwasawa-theoretic Heegner Kolyvagin systems modulo p.","tokens_in":35791,"tokens_out":18754,"duration_ms":191748,"significance":"If the results are correct, the paper is substantial: it gives a new route from a primitivity statement to the full Howard/Perrin-Riou main conjecture, it covers cases of analytic rank greater than one, and it does not require semistability of the elliptic curve. The improved Kolyvagin-system formalism with the explicit error term d(κ) is a natural and useful strengthening of Howard's framework, and the conditional characterization of primitivity in Theorem B is an interesting converse direction. The paper is honest about its hypotheses and about which results are imported from Wei Zhang, Castella-Wan, Cornut-Vatsal, Jetchev-Skinner-Wan, and Castella-Hsieh; I did not detect circular reasoning, since Theorem A does not assume Howard's conjecture and Theorem B states it explicitly as an assumption. The main weaknesses are that several load-bearing local congruences are asserted tersely and that the application of the formal theory to anticyclotomic twists is not fully checked in the text.","major_comments":[{"comment":"The implication 'κ primitive ⇒ κ^Hg primitive' is the crucial bridge in Theorem A, but it is established in two very short paragraphs. The displayed congruences are not derived, and the passage from nonvanishing of D0DmΦP[m] modulo p to a nonzero class in H^1_{FΛ(m)}(K,T/m) is not written out. There is also a concrete inaccuracy: in the inert ordinary case, reducing Φ=(p+1)^2−a_p^2 modulo p gives (1−a_p^2)D0DmP[m], not (a_p^2−1)D0DmP[m] as printed. The sign difference is a p-adic unit and hence harmless for the intended conclusion, but it shows that these computations are delicate. Because a single missing or incorrect p-factor in these congruences would break the chain κ primitive ⇒ κ^Hg primitive ⇒ Λ-primitive ⇒ equality in Howard's Main Conjecture, I ask the authors to give a complete derivation of all three congruences, including the use of D0σ_p=D0σ_p^*=D0, and to explain explicitly why the resulting Kummer class is nonzero modulo p.","section":"§3.2, Propositions 3.2.2 and 3.2.3"},{"comment":"The equality in part (c) when κ is Λ-primitive is a central ingredient of Theorem 3.3.1 and hence of Theorem A. In the proof, the assertion 'Exactly as in [MR04, Lemma 5.3.20], we have κ(P)≠0 when κ is Λ-primitive' is imported without detail, and the subsequent boundedness of d(κ(Q)) as Q varies is the mechanism that upgrades divisibility to equality. Since Definition 2.4.2 gives nonzero images only in quotients of the form T/(P,m^k) and at finite levels L_j, the implication to a nonzero class in the P-completion used in the argument is not immediate. The authors should either prove the relevant analogue of [MR04, Lemma 5.3.20] in Howard's setting or give a precise reference with the hypotheses verified.","section":"§2.4, Theorem 2.4.5"},{"comment":"Theorem B applies Theorem 2.3.6 to the twisted Kolyvagin system κ^χ, but the manuscript does not explicitly check that the Selmer triple for T⊗χ, with χ a nontrivial anticyclotomic character, satisfies the hypotheses (H.0)–(H.5) of Section 2. In particular, the symmetric τ-conjugate self-dual pairing required in (H.4) is not constructed for the twisted module; since the natural Tate pairing pairs T⊗χ with T⊗χ^{-1}, this verification is not automatic. This is load-bearing because Theorem 2.3.6 supplies both the rank-one statement and the error-term formula used in the proof of Theorem B. The authors should state the Selmer triple for T⊗χ explicitly and verify the axioms, or point to a precise reference where this verification is carried out.","section":"§4.4 and §5.2, twisted Kolyvagin systems"}],"minor_comments":[{"comment":"The text contains many typos and formatting errors, including 'HOW ARD'S' and 'KOL YV AGIN' in the title, 'conjeturally', 'endormorphism', 'trivally', 'W e', 'Frobeniuses', and 'a explicit reciprocity law' in the abstract. An editorial pass is needed.","section":"Throughout"},{"comment":"In the analytic rank one branch of Theorem B, the notation 'z_χ := κ^χ_1' is reused with χ trivial, and the displayed application of Corollary 5.1.2 is only valid in the first branch. The text should separate the two cases more clearly, especially because Corollary 5.1.2 requires s>f_0, which fails for trivial χ.","section":"§5.2, proof of Theorem B"},{"comment":"The sign discrepancy in the inert ordinary congruence should be corrected, and the statement should specify that the displayed scalars are only determined up to a p-adic unit, which is all that the primitivity argument needs.","section":"§3.2, Proposition 3.2.2"}],"recommendation":"major_revision","confidential_remarks":"The manuscript appears to be a lightly edited thesis chapter, and the presentation quality is below what one expects in a journal submission. The mathematical core is promising and I see no obvious fatal error, but the referee should be asked to verify the local congruences in Propositions 3.2.2 and 3.2.3 and the application of the formalism to twisted representations in Theorem B before acceptance."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The new thing here is an error term d(kappa) in Howard's Kolyvagin-system formalism, which upgrades the usual divisibility to an equality exactly when the Heegner Kolyvagin system is primitive. That is a real improvement, and Theorem A follows cleanly from it. The formal core in Section 2 is coherent: Theorem 2.3.6 refines Howard's inequality in a natural way, and the adaptation of the Mazur–Rubin argument to get equality from Lambda-primitivity is credible. The paper is also honest—it flags Assumption 1.1.1 as likely removable and does not hide that Theorem B is conditional on Howard's main conjecture.\n\nThe weakest point is the bridge between the two primitivity notions. Propositions 3.2.2 and 3.2.3 assert the mod-p congruences that transport primitivity from the finite-level Kolyvagin system kappa to the Iwasawa-level system kappa^Hg, and the derivations are compressed into single lines with no appendix revisiting them. The stress-test note correctly points out a sign issue in the inert ordinary case: direct reduction of Phi at p gives (1 - a_p^2), not (a_p^2 - 1). That sign is a p-adic unit under (not anom), so it is harmless for primitivity, but it illustrates exactly how easy it is to drop a unit. A missing p-factor anywhere in those three congruences would break Theorem A, because the whole argument reduces to choosing m where the reduced class is nonzero. I do not see an actual error, but this is the spot that needs the most careful checking.\n\nTwo other soft spots, both minor in proportion. Theorem 4.2.1 leans on an internal step of [JSW17] that is not reproduced; that is a verification burden, not an error. Some local computations at primes above p in the control theorem are terse. None of these strike me as load-bearing in the way the congruences are.\n\nWho this is for: Iwasawa theorists working on Heegner points, especially anyone using Howard's formalism. It deserves a serious referee. An expert should spend real time on Propositions 3.2.2/3.2.3 and on the JSW17 application; if those check out, this is accept-level work. I would send it to peer review, not desk reject.","headline":"A serious, well-structured paper that upgrades Howard's divisibility to equality under a primitivity hypothesis; the one load-bearing spot is the terse mod-p congruence bridge, which needs a careful check.","tokens_in":36465,"tokens_out":2122,"would_cite":false,"duration_ms":22583,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["11R23","11G05","11G40"],"pacs":[],"model":"deepseek-v4-flash","headline":"The paper proves that a primitive Heegner-point Kolyvagin system upgrades Howard's divisibility to the full Howard Main Conjecture, and it establishes an equivalence between that conjecture and primitivity plus a Tamagawa-factor condition.","keywords":["Howard's Main Conjecture","Heegner points","Kolyvagin systems","Iwasawa theory","elliptic curves","anticyclotomic extension","p-adic L-functions","primitivity"],"falsifier":"Take an elliptic curve $E/\\mathbb{Q}$, a prime $p\\ge 5$, and a quadratic imaginary field $K$ satisfying the theorem's hypotheses with $\\kappa$ primitive, and compute the mod-$p$ reductions of $\\kappa$ and of the Iwasawa-theoretic class $\\kappa^{Hg}$ at a square-free level where the Selmer group has rank one. The central bridge predicts the exact congruences $D_0D_m\\Phi P[m]\\equiv(a_p-1)^2D_0D_mP[m]\\pmod p$ in the split ordinary case and $D_0D_mP_0[m]\\equiv-2D_0D_mP[m]\\pmod p$ in the supersingular case; exhibiting any curve where one of these fails, or where $\\kappa$ is primitive yet the two characteristic ideals have different $p$-adic valuations, would refute Theorem A.","tokens_in":35215,"feed_emoji":"","tokens_out":18344,"duration_ms":150106,"temperature":0.7,"pith_summary":"The paper's central claim is that Howard's Main Conjecture for Heegner points—an equality of characteristic ideals in anticyclotomic Iwasawa theory—follows from a single explicit condition: the usual finite-level Kolyvagin system $\\kappa$ is primitive, meaning its reduction modulo $p$ is nonzero. The proof works by refining Howard's Kolyvagin-system formalism with an error term $d(\\kappa)$ that measures the gap between a Selmer module's length and the length of the quotient by the Heegner class; this gap vanishes exactly when $\\kappa$ is primitive. The author then shows, through mod-$p$ congruences that require $p$ to be non-anomalous, that primitivity of $\\kappa$ forces the Iwasawa-theoretic Heegner system $\\kappa^{Hg}$ to be $\\Lambda$-primitive, which is precisely what upgrades the known divisibility to equality. The paper also proves a partial converse: assuming Howard's Main Conjecture, $\\kappa$ is primitive exactly when no split prime Tamagawa factor is $p$-divisible, and a full equivalence under Assumptions 1.1.7 and 1.1.1 states that $\\kappa$ is primitive precisely when Howard's Main Conjecture and the Tamagawa condition both hold. If correct, the paper turns a conjecture about $p$-adic $L$-functions and Selmer groups into a statement about indivisibility of Heegner points, extending known cases to analytic rank greater than one and to non-semistable curves.","feed_headline":"Primitive Heegner classes prove Howard's main conjecture","feed_subtitle":"Equality replaces divisibility, covering curves of analytic rank greater than one.","key_machinery":"The engine of the proof is a refinement of the Kolyvagin-system formalism for conjugate-self-dual Galois representations. For a Kolyvagin system $\\kappa$ over a discrete valuation ring, the paper proves the precise error formula $\\operatorname{length}(M)=\\operatorname{length}(H^1_F(K,T)/R\\kappa_1)-d(\\kappa)$, where $M$ is the torsion module in the structure $H^1_F(K,A)\\simeq D\\oplus M\\oplus M$, and $d(\\kappa)$ is an integer attached to $\\kappa$ with $d(\\kappa)=0$ exactly when $\\kappa$ is primitive. This is the object that converts the known inequality into an equality: the divisibility in Howard's Main Conjecture is the inequality side, and the error term vanishes precisely under primitivity. To get from finite-level primitivity to the Iwasawa-theoretic condition, the paper uses the mod-$p$ congruences of Propositions 3.2.2 and 3.2.3, which identify the reduction of the $\\Lambda$-adic Heegner system $\\kappa^{Hg}$ with a nonzero scalar multiple of the reduction of $\\kappa$; in the split ordinary case the scalar is $(a_p-1)^2$ and in the supersingular case it is $-2$, and these are exactly where the non-anomalous hypothesis enters.","core_discovery":"Concretely, under the paper's Assumptions 1.1.1 and 1.1.3, the non-anomalous condition, and primitivity of the Heegner Kolyvagin system $\\kappa$, Theorem A establishes the full statement of Howard's Main Conjecture: the torsion $\\Lambda$-module $M$ satisfies $\\operatorname{char}(M)=\\operatorname{char}(M)^\\iota$, the Selmer group $X$ is pseudo-isomorphic to $\\Lambda\\oplus M\\oplus M$, and $\\operatorname{char}(M)=\\operatorname{char}(H^1_{F_\\Lambda}(K,T)/\\Lambda\\kappa^{Hg}_1)$. In other words, the divisibility obtained in the original and supersingular formulations is upgraded to equality. The paper also proves a converse direction: assuming Howard's Main Conjecture, $\\kappa$ is primitive if and only if no split prime Tamagawa factor is $p$-divisible, and an equivalence theorem states that $\\kappa$ is primitive exactly when Howard's Main Conjecture and that Tamagawa condition both hold. The route combines the refined Kolyvagin-system formalism with an anticyclotomic control theorem and an explicit reciprocity law for Heegner points, which ties the BDP $p$-adic $L$-function main conjecture to the index of the Heegner classes.","pith_inferences":["If primitivity is the sharp hypothesis, the same strategy may prove Howard's Main Conjecture for abelian varieties of $\\mathrm{GL}_2$ type or higher-weight modular forms wherever an explicit reciprocity law is available.","The error term $d(\\kappa)$ may coincide with the classical Kolyvagin-system divisibility exponent in a wider range of Selmer structures; proving equality would make the formalism uniform and give a single primitivity criterion.","Because the proof bypasses semistability, a natural test is to compute both characteristic ideals and the error term on a non-semistable curve of analytic rank 2; a failure would localize the obstruction to the mod-$p$ congruences rather than to the underlying Iwasawa theory.","The equivalence suggests a purely geometric reformulation: the main conjecture may be equivalent to the indivisibility of Heegner points together with the absence of split Tamagawa factors, a statement approachable by congruences between modular forms."],"forward_implications":["Howard's Main Conjecture becomes a theorem in every case where primitivity of the Heegner Kolyvagin system is known, including curves of analytic rank greater than one and curves that are not semistable.","The equality of characteristic ideals $\\operatorname{char}(M)=\\operatorname{char}(H^1_{F_\\Lambda}(K,T)/\\Lambda\\kappa^{Hg}_1)$ follows from the error-term formula once $\\kappa^{Hg}$ is known to be $\\Lambda$-primitive.","Theorem C gives an arithmetic criterion: under the stated hypotheses, Howard's Main Conjecture holds exactly when $\\kappa$ is primitive and no split prime Tamagawa factor is $p$-divisible.","In the supersingular reduction case, the same argument proves the `+`-version of the main conjecture under the analogous hypotheses.","The proof shows that an explicit reciprocity law can be used to read primitivity of $\\kappa$ from the $p$-adic valuation of a BDP $L$-value, connecting the main conjecture to classical Kolyvagin structure theorems."],"supporting_citations":[{"why":"Supplies the original Kolyvagin-system formalism and the inequality that the paper refines with the error term $d(\\kappa)$.","marker":"[How04a]"},{"why":"Formulates Howard's Main Conjecture and establishes the ordinary divisibility that this paper upgrades to equality.","marker":"[How04b]"},{"why":"Provides the classical Kolyvagin-system framework in which an error term converts inequalities into equalities.","marker":"[MR04]"},{"why":"Proves primitivity of the Heegner Kolyvagin system in broad cases, supplying the hypothesis that yields new instances of Howard's Main Conjecture.","marker":"[Zha14]"},{"why":"Constructs the supersingular analogue of the Iwasawa-theoretic Heegner system and defines the '+' Selmer condition used in the supersingular part.","marker":"[CW16]"},{"why":"Guarantees infinitely many anticyclotomic twists with nonzero Heegner base classes, needed for the converse direction.","marker":"[CV07]"},{"why":"Provides the anticyclotomic control theorem used to relate the BDP $p$-adic $L$-function to Selmer groups and special values.","marker":"[JSW17]"},{"why":"Supplies the explicit reciprocity law expressing the BDP $p$-adic $L$-function in terms of the logarithm of the Heegner class, the bridge for Theorem B.","marker":"[CH18]"},{"why":"Establishes the analytic-rank-one case and the equivalence between Howard's and BDP main conjectures, which Theorem C builds on.","marker":"[BCK18]"},{"why":"Provides Kolyvagin's structure theorem relating the divisibility exponent to the index of the Heegner point, used in the analytic-rank-one comparison.","marker":"[Kol91]"}],"fun_headline_variants":["Howard's main conjecture now an equality","Heegner point conjecture upgraded to equality","Kolyvagin system yields Howard's conjecture","Equality replaces divisibility in Howard conjecture","Primitive Heegner classes force Howard's equality"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is the exactness of the mod-$p$ congruences (Propositions 3.2.2 and 3.2.3) that identify the reduction of the Iwasawa-theoretic Heegner system $\\kappa^{Hg}$ with a nonzero scalar multiple of the reduction of the finite-level system $\\kappa$; the paper's derivation of these congruences is terse, and any $p$-factor error there would destroy the bridge from primitivity of $\\kappa$ to $\\Lambda$-primitivity of $\\kappa^{Hg}$.","fun_headline_variants_meta":{"raw":{"variants":["Howard's main conjecture now an equality","Heegner point conjecture upgraded to equality","Kolyvagin system yields Howard's conjecture","Equality replaces divisibility in Howard conjecture","Primitive Heegner classes force Howard's equality"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000254,"raw_usage":{"total_tokens":1541,"prompt_tokens":891,"completion_tokens":650,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":507,"completion_tokens_details":{"reasoning_tokens":582}},"tokens_in":507,"tokens_out":650,"duration_ms":6218,"temperature":1.0,"reasoning_tokens":582,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T11:19:39.085895+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Take an elliptic curve $E/\\mathbb{Q}$, a prime $p\\ge 5$, and a quadratic imaginary field $K$ satisfying the theorem's hypotheses with $\\kappa$ primitive, and compute the mod-$p$ reductions of $\\kappa$ and of the Iwasawa-theoretic class $\\kappa^{Hg}$ at a square-free level where the Selmer group has rank one. The central bridge predicts the exact congruences $D_0D_m\\Phi P[m]\\equiv(a_p-1)^2D_0D_mP[m]\\pmod p$ in the split ordinary case and $D_0D_mP_0[m]\\equiv-2D_0D_mP[m]\\pmod p$ in the supersingular case; exhibiting any curve where one of these fails, or where $\\kappa$ is primitive yet the two characteristic ideals have different $p$-adic valuations, would refute Theorem A.","supporting_citations":[],"review_version":1}