{"id":"37d120a2-cce6-4e14-8ed5-336df242212e","arxiv_id":"2511.08705","paper_version":4,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":2,"one_line_summary":"Every 7-tuple of non-standard subgroups of a finite almost simple group has conjugates whose intersection is trivial — only M24 needs all seven.","lead":"This paper proves a conjecture from the theory of permutation groups: in any finite almost simple group, any seven non-standard subgroups can be conjugated so that their intersection is trivial, with the Mathieu group M24 as the unique exception. It completes a program begun with Burness and uses probabilistic counting, deep structural results on groups of Lie type, and extensive computer verification.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"E8(q) bound in Prop 10.10 rests on an unverified coverage claim for maximals with |H|>q^88, while §7.1 says the S-collection is open.","rationale":"The reader's verdict is CONDITIONAL, and its weakest_assumption identifies the same structural gap: E8(q) non-parabolic fixed point ratios are obtained by inspection of [26, 4.2/4.4–4.6] despite the S-collection being open. My stress-test confirms this is the most load-bearing concern: the entire exceptional-group theorem depends on Prop 10.10, whose proof contains an unproved coverage assertion. The computational layer (unshipped code, private Foulkes data) is secondary; even if all computations were released, the classification assumption would remain. A single independent category check—that every maximal subgroup of order >q^88 is on the known list—settles the matter. Since this is exactly what the CONDITIONAL verdict requires, no change to the reader's verdict is needed.","tokens_in":75341,"tokens_out":5172,"duration_ms":54075,"concrete_test":"Independently verify the [26,4.2] coverage assertion: list all E8(q) maximal subgroups with |H|>q^88 from the published classifications (parabolic, maximal-rank, subfield, Borovik for p>5), and prove any S-class subgroup not in that list has order ≤q^88 using Landazuri–Seitz lower bounds on the minimal degree of a nontrivial irreducible representation of the socle. If a candidate with |H|>q^88 is missing, Prop 10.10 collapses; if every S-subgroup has order ≤q^88, the concern is resolved.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The proof of Theorem 7.1 for E8(q) (Prop 10.10) assumes that all non-parabolic maximal subgroups of order >q^88 are described by [26,4.2] and satisfy the fpr bounds c_i from [26,4.4–4.6]. This is load-bearing because the smallest classes in L_2 have size only ~q^125, so an unlisted H of order, say, q^90 would give fpr ≈ q^{-35} > q^{-37}=c_2. The manuscript itself (§7.1) states that the S-collection of E8 maximal subgroups is not yet classified even up to isomorphism, and no argument is given that any such H must have order ≤q^88. Without this coverage claim the uniform c_i bounds, and therefore the d_i c_i^6 < q^{-1} count, do not follow.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper proves a generalisation of Cameron's base size conjecture to heterogeneous tuples of subgroups. For a finite almost simple group G, it defines R_ns(G) as the least r such that every r-tuple of non-standard maximal subgroups admits a regular orbit on the product of coset spaces, and proves Theorem 2: R_ns(G) ≤ 7, with equality if and only if G = M24. The proof splits into classical and exceptional Lie type cases. For classical groups it proves the stronger Theorem 3.1: R_ns(G) ≤ 5, with equality only for U6(2).2, and P(G,4)→1 as |G|→∞. For exceptional groups it proves Theorem 7.1: every 6-tuple of core-free subgroups of an almost simple exceptional group is regular, with Q(G,τ)→0 as q→∞. The methods are probabilistic, based on fixed point ratio estimates, with extensive use of published classification results, known fixed point ratio bounds, and computational checks in GAP/Magma.","tokens_in":75400,"tokens_out":12460,"duration_ms":122209,"significance":"If correct, this settles a natural and well-motivated generalisation of Cameron's conjecture and gives a remarkably strong uniform bound for non-conjugate tuples. The paper contains useful standalone contributions: the fixed point ratio bounds in Theorem 4.1 and Tables 13–17, the eta-function estimates in Section 5, and the refined asymptotic statement P(G,4),P(G,6)→1. There is no circularity: the target theorem is not used as an input, the constants in the bounds are fixed and derived, and the small-base-size results of [18] are used only as comparisons or for special tuples. However, the E8(q) case rests on an external structural claim about large non-parabolic maximal subgroups that is not reconciled with the paper's own statement that the S-collection of E8(q) is unclassified, and several essential computations rely on private or unshipped data. The classical part and the E1/E2 exceptional parts appear sound; the E8 gap is load-bearing and needs to be closed or made explicit.","major_comments":[{"comment":"For E8(q), the non-parabolic case splits at |H|≤q^88. For |H|>q^88, the proof says: 'The possibilities for H are described in [26, 4.2]' and fpr(x,G/H)≤c_i follows by inspecting [26, 4.4–4.6]. This is load-bearing: the smallest classes in L2 have size about q^125, so an unlisted maximal subgroup of order q^90 would give fpr≈q^{-35} > q^{-37}=c_2. Yet §7.1 states that the S-collection of E8(q) maximal subgroups is not classified even up to isomorphism, and the paper gives no argument that every unclassified member has order ≤q^88. The sentence 'The current state of the art regarding the groups in class S will not be relevant to us' is therefore not justified by the text. Please supply a proof or a precise reference for the coverage claim for maximal subgroups of order >q^88, or restrict/restructure the argument.","section":"§7.1, Prop 10.10 (E8(q))"},{"comment":"Several essential steps rest on computations that cannot be audited from the manuscript. The Foulkes-function data [65] is a private communication, and numerous 'one can check' assertions (e.g., Propositions 5.1–5.3, 6.3–6.5, 10.10) are not accompanied by code or output. The Magma/GAP fusion computations in Propositions 8.2 and 8.4 are described only in outline. These computations cover the small-q cases and the parabolic fixed-point-ratio estimates that are needed for the exceptional-group proof. Please deposit the code and/or output, or provide enough detail (inputs, intermediate data, verification scripts) for an independent reader to reproduce the calculations.","section":"Sections 2.2, 5, 6, 9, 10"}],"minor_comments":[{"comment":"Typos: 'GENERALISA TION' in the title; 'permuatation' in Lemma 2.3; 'the proofs of the bounds in the proofs of' in Proposition 4.12; 'it is left to check' in Proposition 4.4. A careful proofreading pass is needed.","section":"Throughout"},{"comment":"Theorem 7.1 is stated for 6-tuples of core-free subgroups, but the fixed point ratio estimates in the proof are stated for maximal subgroups. It would be helpful to add the standard observation that if H≤M, then fpr(x,G/H)≤fpr(x,G/M), so one may pass to a maximal overgroup; this is implicit but should be explicit.","section":"Theorem 7.1"},{"comment":"Several β-values are given as decimals (0.4, 0.225, 0.32). Exact rationals would be preferable, especially since the proofs refer to inequalities such as log|x^{G0}∩H|/log|x^{G0}|<0.68. ","section":"Table 7"},{"comment":"Many cases in Propositions 4.3–4.6 and 4.10–4.16 are left 'for the reader to check'. This is acceptable in a research paper, but the overall verification burden is large; a supplementary file with the omitted computations would greatly improve reproducibility.","section":"Section 4"}],"recommendation":"major_revision","confidential_remarks":"This is a strong and substantial paper. The classical-group part and the E1/E2 exceptional-group parts seem convincing, and the paper contains valuable technical results. My main concern is the E8(q) coverage claim in Prop 10.10, which is load-bearing and appears inconsistent with the stated openness of the S-collection. The second concern is computational reproducibility. If the authors can supply a precise reference or proof for the |H|>q^88 classification used for E8(q), and make the computational data available, I would support publication."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The headline: this paper proves the natural generalisation of Cameron's base size conjecture – for any almost simple G, any 7-tuple of non-standard core-free maximal subgroups has a regular orbit, with equality iff G = M24, plus sharper bounds (R_ns ≤ 5 for classical, ≤ 6 for exceptional). If it holds up, it closes a program that Burness et al. started. The Lie type cases are genuinely new: dropping the conjugacy condition forces uniform worst-case fixed point ratio bounds, and the paper supplies those (Theorem 4.1 for low-dimensional classical groups, Propositions 8.1–8.4 for exceptional non-parabolic actions). The asymptotic corollary (Theorem 4) is a nice bonus.\n\nThe structure is sound. Lemma 2.1 is the standard union bound; the classical/exceptional split is clean; the fixed point ratio estimates are mostly proven or verified by explicit computation. I don't see circularity: no fitted parameters, the target theorem never enters as an input, and the self-citation [2] covers only alternating/sporadic socles, which is legitimate.\n\nSoft spots, in order of how much they bother me:\n\n1. The verification layer. Many claims are 'one can check' or 'using Magma one verifies', and the code and data are not shipped. Luebeck's Foulkes functions are private communication. For a proof of this scale that is normal, but it needs a companion artifact – scripts with a commit hash, data files – before I would call it fully verified. The E6(2) P1/P6 case is described as 'random search'; that needs a deterministic certificate.\n\n2. The E8(q) non-parabolic bound (Prop 10.10) leans on [26, 4.2] for the claim that any maximal subgroup of E8(q) of order > q^88 is one of the listed known types. Since the S-collection is still open (as the paper notes in §7.1), a referee should check that [26, 4.2] indeed rules out unclassified subgroups of that size. My reading is that it should, because the unknown S-subgroups are small (order at most roughly q^80), which is why the author says they are irrelevant. But the paper does not spell that out; it just cites. That is fine in principle, but it is a load-bearing citation and deserves a careful look.\n\nThis is a strong paper. The math is serious, the case analysis is thorough, and the main theorem is a genuine completion. I would send it to a good referee rather than desk reject.","headline":"Proves the regularity-number analogue of Cameron's conjecture (R_ns(G) ≤ 7, equality iff M24) with new Lie type bounds; substantial and probably correct, but the computational layer needs to be released and the E8(q) coverage citation checked.","tokens_in":76127,"tokens_out":10273,"would_cite":true,"duration_ms":96145,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["20B05","20B15","20D06"],"pacs":[],"model":"deepseek-v4-flash","headline":"Seven conjugates always suffice for non-standard subgroup tuples","keywords":["base size","regularity number","non-standard maximal subgroups","fixed point ratios","almost simple groups","groups of Lie type","probabilistic group theory","sporadic group M24"],"falsifier":"Produce a maximal subgroup of E8(q) with order larger than q^88 that is not among the families cited in the paper, and exhibit a prime-order element whose fixed-point proportion exceeds the recorded bounds; this would collapse the uniform estimate behind the exceptional-group theorem.","tokens_in":75003,"feed_emoji":"🧮","tokens_out":7639,"duration_ms":83749,"temperature":0.7,"pith_summary":"This paper proves that a single absolute bound governs how many subgroups of a finite almost simple group must be conjugated before they can be made to meet trivially, even when the subgroups are chosen independently rather than as repeats of one stabilizer. The main theorem states that any seven non-standard maximal subgroups can be simultaneously conjugated to have trivial intersection, and that seven is needed only by the sporadic group M24 with its M23 stabilizer. For classical groups five conjugates suffice, with the lone exception U6(2).2; for exceptional groups of Lie type, any six proper subgroups suffice. On the way the paper establishes quantitative probabilistic estimates—fixed-point-ratio bounds and zeta-type function evaluations—that are likely to be reused independently.","feed_headline":"Seven conjugates always meet trivially in almost simple groups","feed_subtitle":"A 1990s base-size conjecture holds even when the subgroups are different; classical groups need only five","key_machinery":"The engine is the probabilistic method: for a tuple τ, the probability that a random point lies outside all regular orbits is bounded above by Σ_i |x_i^G| · Π_j fpr(x_i, G/H_j), where x_i runs over prime-order conjugacy classes and fpr denotes the fixed point ratio. The proof reduces to worst-case fixed-point-ratio estimates over all non-standard subgroups together with the zeta-type function η_G(t) = Σ_C |C|^{-t} over prime-order G0-classes; showing η_G(t) < 1 for a suitable t forces the tuple to be regular. Low-dimensional classical groups require sharpened fixed-point-ratio bounds derived from maximal-subgroup data and conjugacy-class tables; high-dimensional classical groups are handled","core_discovery":"The central discovery is that the base-size bound seven survives a substantial weakening of the old conjectures: the tuple of subgroups is no longer required to be conjugate. Defining the regularity number R_ns(G) as the least k such that every k-tuple of non-standard core-free maximal subgroups is regular—that is, admits an orbit with trivial pointwise stabilizer on the product of coset spaces—the paper shows R_ns(G) ≤ 7 for every finite almost simple group, with equality exactly for G = M24. The classical-group case is settled more sharply at R_ns(G) ≤ 5, with equality iff G = U6(2).2, and the exceptional-group case at R(G) ≤ 6, meaning in particular that six arbitrary proper subgroups of","pith_inferences":["The domination argument—any core-free subgroup sits inside a maximal one—suggests that the 7-tuple statement may extend to arbitrary non-standard core-free subgroups, not only maximal ones, although the paper only formulates the maximal version.","The refined zeta-type bounds established here, such as η_G(1/5) < 0.9 for PSp6(q) and Ω7(q), are standalone quantitative estimates that could be reused in spectral or random-walk questions on finite groups of Lie type.","A direct computational check of every 6-tuple among the known maximal subgroups of E8(2) would test the sharpest finite case of the exceptional-group theorem; a non-regular 6-tuple there would refute it, while a positive check would still leave the infinite family dependent on the open classification.","If a future classification of the missing E8(q) maximal subgroups produces a previously unknown family, the fixed-point-ratio bounds behind the exceptional-group theorem would need to be re-verified against that family—a concrete revision point that would not necessarily change the main theorem."],"forward_implications":["For every finite almost simple group, any seven non-standard maximal subgroups can be conjugated so that their intersection is trivial; only M24 with M23 stabilizers genuinely requires seven.","In classical groups, five conjugates always suffice, with the single exception U6(2).2, whose unique non-regular 4-tuple is four copies of U4(3).2².","In every simple exceptional group of Lie type, any six proper subgroups can be simultaneously conjugated to have trivial intersection.","As |G| grows, the probability that a randomly chosen 6-tuple of non-standard subgroups is regular tends to 1.","The original single-action conjecture is recovered as the special case where all subgroups in the tuple are pairwise conjugate."],"fun_headline_variants":["Cameron's base-size bound holds for non-conjugate subgroups","Seven non-standard subgroups always yield a regular orbit","Classical groups need only five in generalized Cameron result","Exceptional groups: six arbitrary subgroups meet trivially"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The load-bearing premise is that, in the unresolved E8(q) case, the worst-case fixed point ratio is achieved by a maximal subgroup from the already classified families; the full classification of E8(q) maximal subgroups is still open.","fun_headline_variants_meta":{"raw":{"variants":["Cameron's base-size bound holds for non-conjugate subgroups","Seven non-standard subgroups always yield a regular orbit","Classical groups need only five in generalized Cameron result","Exceptional groups: six arbitrary subgroups meet trivially"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000431,"raw_usage":{"total_tokens":2155,"prompt_tokens":977,"completion_tokens":1178,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":721,"completion_tokens_details":{"reasoning_tokens":1112}},"tokens_in":721,"tokens_out":1178,"duration_ms":11894,"temperature":1.0,"reasoning_tokens":1112,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-03T22:47:47.751018+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Produce a maximal subgroup of E8(q) with order larger than q^88 that is not among the families cited in the paper, and exhibit a prime-order element whose fixed-point proportion exceeds the recorded bounds; this would collapse the uniform estimate behind the exceptional-group theorem.","supporting_citations":[],"review_version":1}