{"id":"7b20a45d-e12e-4bec-a0ff-2ac16658440e","arxiv_id":"1908.09512","paper_version":2,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":8.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"The authors prove Perrin-Riou's Heegner point main conjecture for elliptic curves over the rationals under mild hypotheses, and also prove the Iwasawa-Greenberg main conjecture for Bertolini-Darmon-Prasanna p-adic L-functions.","lead":"This paper proves a famous 1987 conjecture by Perrin-Riou about how certain arithmetic data attached to elliptic curves grows along an infinite tower of number fields. It is a major advance in number theory, combining two powerful methods that were previously used separately.","discovery_kind":"unification","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The proof of Theorem A is conditional on the non-anomalous hypothesis through Prop. 3.7's p-adic multiplier; anomalous higher-rank primes remain outside the stated theorem.","rationale":"The reader's weakest assumption identifies the non-anomalous hypothesis, and this is indeed the point where the proof of Theorem A is most exposed. Proposition 3.7 is the linchpin connecting Wei Zhang's results to Howard's equality criterion, and its proof is compressed: the r=1 base case imports the nonvanishing of L_alg(g/K) from [Zha14, Thm. 7.2] and the interpolation formula from [CH18b, Thm. A], while the induction step cites the argument of [Zha14, Thm. 9.1]. The non-anomalous hypothesis is used specifically to ensure e_p(g,1) is nonzero modulo wp, and the authors' Remark A.7 concedes that removing it would require additional arguments involving characters and indices in E(K_n)_chi, not provided in the paper. This is a genuine limitation of the theorem's scope, but it is not a flaw in the stated theorem, since Theorem A assumes non-anomalous. The conditional verdict remains appropriate: the proof structure is coherent, and the main unresolved question is whether the non-anomalous condition can be eliminated in higher rank, as the authors suggest. No evidence of fraud or self-deception was found, and the deferred references are to serious works in the field.","tokens_in":21667,"tokens_out":20059,"duration_ms":212723,"concrete_test":"Re-derive the interpolation identity used in Proposition 3.7 for a concrete anomalous ordinary prime p (e.g., an elliptic curve E/Q with a_p congruent to 1 mod p and p split in K) and a rank-one K satisfying (Heeg), (disc), and Hypothesis ♠. Compute the augmentation of lambda_1(q)^2 modulo wp via [CH18b, Thm. A] and compare it with e_p(g,1) * L_alg(g/K). If the product vanishes while L_alg(g/K) is nonzero modulo wp, the non-anomalous hypothesis is essential to the argument as written; if the T-coefficient of lambda_1(q) modulo wp is nonzero, then the assumption is removable and Proposition 3.7 can be strengthened.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The decisive step in the proof of Theorem A is Proposition 3.7, which asserts that the bipartite Euler system's lambda-system has nonzero image in Lambda/(wp) and thereby triggers Howard's equality criterion (Theorem 3.4, via Lemma 3.6). The proof of Proposition 3.7 rests on showing, for a level-raising form g, that the augmentation of lambda_1(q)^2 is congruent to e_p(g,1) * L_alg(g/K) mod wp, up to a p-adic unit. The factor e_p(g,1) is nonzero only under the non-anomalous hypothesis: a_p not congruent to 1 mod wp when p splits in K, and a_p^2 not congruent to 1 mod wp when p is inert. For the induction step r >= 3, the argument descends to a form of Selmer rank r-2 by level raising as in [Zha14, Thm. 9.1], and the same multiplier condition is inherited at every stage. Thus, when p is anomalous, the displayed nonvanishing can fail, and the proof of Theorem A as written does not apply. This is not an internal inconsistency, because Theorem A explicitly assumes p is non-anomalous; however, Appendix A and Remark A.7 only remove this assumption in the split, analytic-rank-one case, not in the higher-rank cases. Consequently, the paper does not establish the Heegner point main conjecture for anomalous primes, and the claimed equality in Theorem 3.2 is conditional on an arithmetic condition that the authors themselves indicate should be removable.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper proves Perrin-Riou's Heegner point main conjecture for an elliptic curve E/Q with good ordinary reduction at p > 3 and an imaginary quadratic field K satisfying the generalized Heegner hypothesis, under the assumptions that Hypothesis ♠ holds, the residual representation ρ is surjective, and p is non-anomalous. The proof combines Howard's theory of bipartite Euler systems with Wei Zhang's proof of Kolyvagin's conjecture; the key step (Proposition 3.7) is to show that the associated λ-system has nonzero image modulo ℘, which triggers Howard's equality criterion. When p splits in K, the authors also derive the Iwasawa–Greenberg main conjecture for the Bertolini–Darmon–Prasanna p-adic L-function, via an explicit reciprocity law (Theorem 4.4) and an equivalence between the two main conjectures (Theorem 5.2). An appendix gives a rank-one proof that avoids the non-anomalous hypothesis in the split, analytic-rank-one case.","tokens_in":21943,"tokens_out":3698,"duration_ms":38617,"significance":"If correct, this is a major advance in anticyclotomic Iwasawa theory: it establishes the Heegner point main conjecture for a broad class of elliptic curves, including cases where N^- = 1, where N is not squarefree, and where p is inert in K, all of which were previously inaccessible. The paper also supplies the missing details in the literature for the explicit reciprocity law for N^- ≠ 1, and shows the equivalence between the Heegner point and Iwasawa–Greenberg main conjectures in the split case. The argument is a substantial synthesis of deep external results (Howard, W. Zhang, Cornut–Vatsal, Chida–Hsieh, Manning–Shotton), and the paper is careful to state its hypotheses and to identify the role of each input. The proof is not circular: it derives the conjecture from Howard's machinery and W. Zhang's theorem, and the use of the authors' earlier work [CH18a] is a legitimate prior result, not an assumption of the conjecture.","major_comments":[{"comment":"The proof of Proposition 3.7 is not written in full at the load-bearing step. The case r = 1 is explained, but for r ≥ 3 the argument is summarized as 'by the argument in the proof of Theorem 9.1 we can find a form g2 ... with associated Selmer rank equal to r − 2', and then the induction hypothesis is applied without verifying that the p-adic multiplier e_p(g,1) remains nonzero at every descent stage. Since the nonvanishing of λ modulo ℘ is exactly the input that triggers Howard's equality criterion in Theorem 3.4, this is a central point. The paper's own Remark A.7 and Appendix A show that the non-anomalous assumption can be removed only in the split analytic-rank-one case, so Theorem A remains conditional on an arithmetic condition whose scope in higher rank is not resolved. This is not an internal inconsistency, but the main theorem's proof needs a complete induction argument or an explicit statement that the higher-rank case is deferred.","section":"§3, Proposition 3.7"},{"comment":"The explicit reciprocity law for N^- ≠ 1 is a key ingredient in the proof of Theorem B, and the introduction notes that the details for this extension were previously missing in the literature. However, the proof given here says only that after replacing Proposition 3.3 by Proposition 4.1, 'the argument in [CH18a, Thm. 5.7] applies verbatim'. The paper should provide a full verification of the Shimura-curve-to-Igusa comparison and of the formal-group logarithm computation in the general N^- case, rather than leaving this to a verbatim-reference claim. Without these details, the equivalence in Theorem 5.2 is not self-contained.","section":"§4.2, Theorem 4.4"},{"comment":"The construction of the bipartite Euler system (κ, λ) is summarized via references to Chida–Hsieh and Bertolini–Darmon, but the verification that the hypotheses of Howard's theory, such as [How06, Hyp. 2.2.4 and 2.3.1], hold in this setting is not carried out. Lemma 3.6 asserts that Hypothesis 2.3.1 holds by [BD05, Thm. 3.2], and the existence of core vertices is invoked from [How06, Cor. 2.4.9], but the reader cannot check from the present text that all technical conditions (e.g., the admissibility index and the ordinary local conditions at primes dividing N^- m) are satisfied uniformly for the systems constructed in Theorem 3.3. This is a necessary part of the application of Howard's criterion and should be spelled out.","section":"§3, Theorem 3.3 and Lemma 3.6"}],"minor_comments":[{"comment":"The phrase 'f is assumed to be ordinarity at p' should read 'f is assumed to be ordinary at p'.","section":"§2, p. 5"},{"comment":"The sentence 'the image of λ_1(q)^2 under the augmentation map ... is given by e_p(g,1)·L_alg(g/K) (mod ℘) up to a p-adic unit' should specify the normalization of the isomorphism H^1_ord(K_q, T_j) ≅ Λ/℘^jΛ, since a different normalization changes the unit but not the nonvanishing claim; this would help readers verify the congruence.","section":"§3, Proposition 3.7"},{"comment":"The nonvanishing of loc_p(κ∞) is derived from the implication κ_χ ≠ 0 ⇒ loc_v(κ_χ) ≠ 0, but the cited [Nek07, Thm. 3.2] is used in a way that should be briefly justified: namely, that the Heegner-class construction gives a point whose residue class is nonzero when the global class is nonzero. A short explanation would improve readability.","section":"§4.2, Corollary 4.5"},{"comment":"The phrase 'under mild hypotheses' in the abstract and introduction undersells the non-anomalous condition, which is a substantive arithmetic hypothesis on (E,p,K) and is not removed in the main theorem except in the special case of Appendix A; consider rephrasing to 'under explicit hypotheses'.","section":"§1, Theorem A"}],"recommendation":"major_revision","confidential_remarks":"The paper is clearly a serious and substantial contribution, and the overall strategy is convincing. My main reservation is that the proof of Proposition 3.7, which is the pivotal nonvanishing input, is sketched rather than fully written, and the non-anomalous hypothesis is not addressed beyond a special case. I would be willing to accept the paper after a revision that expands the induction step, fully verifies the hypotheses of Howard's framework, and gives a complete proof of Theorem 4.4 for N^- ≠ 1. The paper's reliance on [CH18a] by one of the authors is legitimate, but the missing details are exactly those flagged in the text, so they should be included rather than deferred."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"You should know this paper before the next seminar. It is the real thing: Burungale, Castella and Kim prove Perrin-Riou's Heegner point main conjecture for elliptic curves, with the old restrictions (squarefree conductor, p split, N^- = 1) gone. They also get the Iwasawa-Greenberg main conjecture for the BDP p-adic L-function in the split case, and an equivalence between the two conjectures. The machinery is Howard's bipartite Euler systems plus Wei Zhang's proof of Kolyvagin's conjecture; the new input is a 'primitivity implies Lambda-primitivity' step, plus an extension of the CH18a explicit reciprocity law to N^- != 1 that was missing in the literature. That last piece is useful in itself.\n\nWhere does the proof stand? The main theorem is conditional on p being non-anomalous (no prime above p divides |~E(F_w)|). That condition enters at Proposition 3.7, where the p-adic multiplier e_p(g,1) must be nonzero to show the lambda-system has nonzero image mod wp and trigger Howard's equality criterion. The authors say the condition should be removable, and Appendix A removes it only in the split, analytic-rank-one case; the higher-rank cases are not covered. So the paper as written does not prove the conjecture for anomalous primes. That is a real limitation, but it is stated plainly and it is not a hidden circularity—the proof is built on independent results.\n\nThe soft spots are mostly about how much is deferred. Several key steps are 'applies verbatim' from Zha14, How06, and MS20; the induction in Proposition 3.7 is sketched. For a paper at this level that is normal, but a referee should ask for the details of Prop 3.7 and the level-raising descent. The reliance on the Manning-Shotton Ihara lemma (to appear) is another thing a referee should keep in mind.\n\nThe citation pattern is healthy: the self-citations are to CH18a, the paper being extended, and that is legitimate. No sign of circularity.\n\nBottom line: if the deferred arguments check out, this is a major result. It deserves a serious referee. Send it to peer review rather than desk reject. For my own work, I would cite it, with the non-anomalous caveat noted.","headline":"A serious proof of Perrin-Riou's Heegner point main conjecture that removes old restrictions; the stated theorem is conditional on non-anomalous p, a limitation the authors themselves flag as removable.","tokens_in":22526,"tokens_out":2058,"would_cite":true,"duration_ms":21173,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["11R23","11F33"],"pacs":[],"model":"deepseek-v4-flash","headline":"The Heegner point main conjecture is proved under mild hypotheses.","keywords":["Iwasawa theory","Heegner points","bipartite Euler systems","anticyclotomic Z_p-extensions","p-adic L-functions","elliptic curves","level raising"],"falsifier":"For a chosen triple satisfying every hypothesis, compute at the trivial character the interpolation formula $\\lambda_1(q)^2 \\equiv e_p(g,1)\\,L_{\\mathrm{alg}}(g/K)$ modulo $\\wp$ for a level-raising prime $q$; the theorem predicts the right-hand side is a $p$-adic unit for some $q$. Finding a triple where the right-hand side vanishes modulo $\\wp$ for every admissible $q$, while the $\\wp$-Selmer rank is at least three, would break the proposition that gives the nonzero image, and with it the proof of the main theorem.","tokens_in":21432,"feed_emoji":"📐","tokens_out":11587,"duration_ms":105837,"temperature":0.7,"pith_summary":"The paper proves the Heegner point main conjecture, an Iwasawa-theoretic statement formulated in 1987 that predicts the size of an elliptic curve's Tate--Shafarevich group over the anticyclotomic $\\mathbb{Z}_p$-extension of an imaginary quadratic field is governed by a system of Heegner points. The main theorem establishes this equality of characteristic ideals for every elliptic curve over $\\mathbb{Q}$ with good ordinary reduction at a prime $p>3$, for every imaginary quadratic field satisfying the Heegner hypothesis, provided a mild ramification condition, surjectivity of the mod-$p$ Galois representation, and a non-anomaly condition on $p$ hold. It removes two restrictions of earlier work: the conductor need not be squarefree and $p$ may be inert in $K$. A second theorem, in the split case, derives the Iwasawa--Greenberg main conjecture for the anticyclotomic $p$-adic $L$-function associated to the curve, so the two conjectures are proved to be equivalent forms of one statement.","feed_headline":"Heegner point main conjecture proved under mild hypotheses","feed_subtitle":"Characteristic-ideal equality now covers inert primes and non-squarefree conductors.","key_machinery":"The load-bearing object is a bipartite Euler system: a pair of compatible families $\\kappa_j(m)$ and $\\lambda_j(m)$ indexed by squarefree products of admissible primes (primes inert in $K$ and satisfying a congruence condition), obtained by level-raising the $p$-stabilized newform at those primes. The $\\lambda$-side consists of elements in the Iwasawa algebra modulo $\\wp^j$, and a criterion from the bipartite Euler system method says the main conjecture's divisibility becomes equality if the $\\lambda$-system has nonzero image modulo $\\wp$. The paper proves this nonzero image by induction on the Selmer rank: level-raising at two admissible primes lowers the rank by two, and in the rank-one base case the non-anomalous hypothesis makes a certain $p$-adic multiplier nonzero, so the interpolated algebraic $L$-value detects the system. In short, the proof upgrades primitivity of the Heegner-point system to the $\\Lambda$-level primitivity that forces equality of characteristic ideals.","core_discovery":"Theorem A is the central claim: under the stated hypotheses, the Pontryagin dual of the Selmer group and the projective limit of $p$-adic Selmer groups both have $\\Lambda$-rank one, and the torsion module $M_\\infty$ satisfies $\\operatorname{Char}_\\Lambda(M_\\infty)=\\operatorname{Char}_\\Lambda(S_\\infty/\\Lambda\\kappa_\\infty)$ together with the self-duality $\\operatorname{Char}_\\Lambda(M_\\infty)=\\operatorname{Char}_\\Lambda(M_\\infty)^\\iota$. In concrete terms, the $p$-primary Tate--Shafarevich group over the tower is measured exactly, up to pseudo-isomorphism, by the Iwasawa module generated by Heegner points. The theorem also yields a $p$-converse: if the Selmer group has corank one, the analytic rank of $L(E/K,s)$ is one. When $p$ splits in $K$, the same proof establishes that the square of the relevant $p$-adic $L$-function generates the characteristic ideal of a modified Selmer group, i.e. the Iwasawa--Greenberg main conjecture.","pith_inferences":["Extension: the non-anomalous hypothesis appears removable, since the rank-one appendix already proves the same conclusions without it; the same induction should yield the main conjecture under the ramification and surjectivity hypotheses alone if the multiplier obstruction can be bypassed.","Extension: the primitivity mechanism is not tied to this particular curve; any setting in which a bipartite Euler system is built from level-raising and a nonvanishing $p$-adic $L$-value could receive the same 'primitivity implies equality' upgrade.","Extension: a concrete numerical check in a small-conductor example could test the mechanism independently of full characteristic-ideal computations, since the theory predicts that the rank-one $\\lambda$-system has a $p$-adic unit image."],"forward_implications":["The Heegner point main conjecture now holds for all triples satisfying the stated hypotheses, including the previously excluded cases $N^-=1$ and $p$ inert in $K$.","The $p$-converse to the analytic-rank-one theorem follows without invoking the structure theorem for Selmer groups: corank-one Selmer implies $L(E/K,s)$ has a simple zero at $s=1$.","In the split case, the Iwasawa--Greenberg main conjecture holds for the anticyclotomic $p$-adic $L$-function: its square generates the characteristic ideal of the dual of the modified Selmer group.","The two main conjectures are equivalent under the stated hypotheses, so any future control theorem or divisibility for one transfers to the other.","The divisibility obtained earlier by the bipartite Euler system method is upgraded to an equality, so the upper and lower bounds on Selmer growth coincide."],"supporting_citations":[{"why":"States the Heegner point main conjecture that the paper proves.","marker":"[PR87]"},{"why":"Extends the conjecture to the case $N^-\\neq 1$ and reduces it to a divisibility via Euler-system machinery.","marker":"[How04b]"},{"why":"Develops bipartite Euler systems and the equality criterion used to upgrade divisibility to equality.","marker":"[How06]"},{"why":"Constructs the bipartite Euler system from level-raised modular forms at admissible primes.","marker":"[BD05]"},{"why":"Refines the construction with the local conditions needed for the Selmer groups.","marker":"[PW11]"},{"why":"Supplies the level-raising and explicit reciprocity laws that define the kappa and lambda systems.","marker":"[CH15]"},{"why":"Provides the nonvanishing and Selmer-rank input behind the proof that the lambda system has nonzero image.","marker":"[Zha14]"},{"why":"Proves the Heegner class $\\kappa_\\infty$ is nontrivial, used for the rank-one statement.","marker":"[CV07]"},{"why":"Provides the explicit reciprocity law for the $p$-adic $L$-function that is extended in Section 4.","marker":"[CH18a]"},{"why":"Constructs the anticyclotomic $p$-adic $L$-function whose square is handled in the split case.","marker":"[BDP13]"}],"fun_headline_variants":["Perrin-Riou Heegner main conjecture proved","Heegner point Iwasawa conjecture: proof","Perrin-Riou's conjecture for Heegner points solved","Heegner points prove anticyclotomic main conjecture"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The argument depends on $p$ being non-anomalous for $E$ over $K$: $p$ divides neither $|\\tilde E(\\mathbb{F}_w)|$ for either prime $w$ of $K$ above $p$. This is what makes the $p$-adic multiplier $e_p(g,1)$ nonzero, and without it the proof that the $\\lambda$-system has nonzero image modulo $\\wp$ does not go through.","fun_headline_variants_meta":{"raw":{"variants":["Perrin-Riou Heegner main conjecture proved","Heegner point Iwasawa conjecture: proof","Perrin-Riou's conjecture for Heegner points solved","Heegner points prove anticyclotomic main conjecture"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000244,"raw_usage":{"total_tokens":1526,"prompt_tokens":936,"completion_tokens":590,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":552,"completion_tokens_details":{"reasoning_tokens":523}},"tokens_in":552,"tokens_out":590,"duration_ms":6010,"temperature":1.0,"reasoning_tokens":523,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T11:09:42.670387+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"For a chosen triple satisfying every hypothesis, compute at the trivial character the interpolation formula $\\lambda_1(q)^2 \\equiv e_p(g,1)\\,L_{\\mathrm{alg}}(g/K)$ modulo $\\wp$ for a level-raising prime $q$; the theorem predicts the right-hand side is a $p$-adic unit for some $q$. Finding a triple where the right-hand side vanishes modulo $\\wp$ for every admissible $q$, while the $\\wp$-Selmer rank is at least three, would break the proposition that gives the nonzero image, and with it the proof of the main theorem.","supporting_citations":[],"review_version":1}