{"id":"186f20e7-0245-4bb6-a837-782cc2588bf1","arxiv_id":"2505.04985","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"For flag-transitive point-primitive 2-designs with λ prime, exceptional simple socles give exactly the Suzuki-Tits ovoid design and a new G2(q) coset-geometry family, while sporadic socles give only three known designs.","lead":"This paper classifies the 2-designs (combinatorial block designs) that admit flag-transitive, point-primitive automorphism groups whose socle is a finite exceptional or sporadic simple group, under the restriction that λ is prime. The result includes one genuinely new infinite family of designs, built from the group G2(q), and shows the sporadic cases reduce to three previously known designs.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Sporadic classification rests on an unshipped GAP enumeration; the final 'ndes' checks on the five surviving K-orbits are the least independently verifiable step.","rationale":"The strongest claim of the paper is a classification theorem: Theorem 1.2 asserts that Table 1 is complete for sporadic socles. Unlike the G2(q) construction in Section 2, which is supported by explicit coset-geometry arguments (Lemmas 2.2–2.4) and by the internal consistency of the parameter equations, the sporadic part is a finite enumeration performed inside GAP. I checked the parameter sets in Table 4 and Table 6: all satisfy the necessary divisibility identities r(k−1)=λ(v−1) and rv=bk, so the candidate list is not empty by accident; the decisive negative information is computational. The reader's weakest assumption names exactly this: the correctness/completeness of the GAP enumeration. I agree. The most load-bearing point is the final 'ndes' classification of B1–B5, because those five tuples are the only ones that pass all structural filters, and the reported failure is not reproducible without the code. I did not find an internal inconsistency in the exceptional-group part; the proof that λ=q+1 is Fermat and the uniqueness of the coset geometry are coherent, provided the cited facts from [24], [13] and [30] hold. The only reservations there (e.g., the asserted normalization of field automorphisms in Theorem 1.1) are minor and do not affect the verdict. The appropriate verdict remains CONDITIONAL, unchanged from the reader.","tokens_in":24322,"tokens_out":33406,"duration_ms":307486,"concrete_test":"Independently re-derive the five negative orbit tests. In GAP or Magma, construct the relevant primitive permutation representation for each surviving row: (i) M23 on 253 points with H=PSL3(4).2^2 (AtlasRep nr 2) and K=2^3:PSL3(2), testing B1, B2, B3; (ii) J2 on 100 points with H=PSU3(3) and K=(([2^6]:3):2):3, testing B4; (iii) J2:2 on 100 points with the corresponding K and B5. For each, form the G-orbit of the listed 7- or 12-subset in the power set and test whether every pair of points lies in exactly λ blocks (5 or 7) by direct pair counting. If any of the five orbits is a base block for a 2-(v,k,λ) design, Theorem 1.2 as stated is false; if all five fail, the final 'ndes' step is confirmed. A public script that also regenerates Tables 4–6 would additionally settle the completeness of the earlier filters.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The completeness part of Theorem 1.2 depends on a GAP 4.12.2 computation whose code and logs are not included. After the elementary filters of Section 3 (divisibility, r | λe, and index divisibility by a maximal subgroup N), 166 candidate tuples remain in Table 6. For each, the authors search for a subgroup K of index b, check K ≤ N and the existence of L ≤ K of index k, and then inspect K-orbits on the v points. This produces exactly five surviving block orbits, B1–B5, each reported as 'ndes' after a Design-package call (Table 6, last column). Every candidate that reaches this stage satisfies all necessary arithmetic conditions, so the negative verdict on B1–B5 is the only mathematical fact excluding those parameter sets. An error in the AtlasRep labels used to identify H or N, in the subgroup-index search (e.g., a missing conjugacy class of subgroups K or L), or in the Design-package check would add a row to Table 1 and falsify Theorem 1.2. The earlier filters (subdegree computations, nsubG/nsubN/nsubK) are equally unverified. No code, random seeds, or detailed logs are shipped, so the computation cannot be reproduced from the manuscript alone.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies flag-transitive and point-primitive automorphism groups of nontrivial 2-(v,k,λ) designs with λ prime, in the almost simple case where the socle is an exceptional group of Lie type or a sporadic simple group. For exceptional socle, it claims that the only examples are the Suzuki-Tits ovoid design and a new infinite family of designs arising from a coset geometry Cos(G2(q), SU3(q):Z2, [q^6]:Z_{q-1}), with q even and q+1 a Fermat prime. For sporadic socle, it claims that only the three parameter sets listed in Table 1 occur, with the proof resting on a GAP-based enumeration in Section 3. The paper also argues that the new G2(q) design is unique up to isomorphism and that the full almost simple group acts on it.","tokens_in":24603,"tokens_out":29260,"duration_ms":277791,"significance":"If the classification is correct, it substantially advances the program of classifying flag-transitive 2-designs with prime λ, and the new G2(q) family is an interesting explicit construction. The paper makes good use of prior reductions and of standard subgroup and orbit information, and the coset-geometry construction is concrete enough to check parameter-by-parameter. The sporadic part, however, is a computer-assisted classification in which the computational evidence is not shipped; this is an important reproducibility weakness. The Suzuki-Tits parameter set also contains a basic design-theoretic inconsistency that must be fixed before the theorem can be accepted.","major_comments":[{"comment":"The displayed parameter set for the Suzuki-Tits ovoid design cannot be correct. For a 2-(v,k,λ) design the identity b=vr/k is forced; substituting (v,b,r,k,λ)=(q^2+1, q^2(q^2+1)/(q−1), q^2, q, q−1) gives vr/k = q(q^2+1), not q^2(q^2+1)/(q−1). Moreover, for q=8, which is admissible because q−1=7 is a Mersenne prime, the printed b is not an integer. The parameter set should be corrected (presumably b=q(q^2+1)) and the citation to [39] checked against the corrected formula.","section":"Theorem 1.1(a) and Abstract"},{"comment":"The inference \"If λ ≠ q+1, then SL2(q) ≤ T_B/R\" is not a consequence of the displayed order |T_B/R| = f1 q(q^2−1)/λ. For q=8, λ=7 and f1=1 this order is 72, whereas |SL2(8)|=504, so the subgroup cannot contain SL2(8). Thus the argument as written does not rule out the possibility λ | q−1. If an additional condition from [39, Lemma 3.8] excludes that possibility, it must be stated; this step is load-bearing because it is what forces λ=q+1.","section":"Lemma 2.1, around Eq. (2.1)"},{"comment":"The case T ≠ G2(q) is dispatched by reference to [39], which by its title concerns non-symmetric designs. The theorem as stated covers all nontrivial 2-designs, including symmetric ones. Please state explicitly how symmetric designs with exceptional socle are excluded (for example by [5], [6] or [11]), or restrict the statement accordingly.","section":"Proof of Theorem 1.1, first sentence"},{"comment":"The completeness part of Theorem 1.2 depends on a GAP enumeration whose code and logs are not included. The entries 'nsubG', 'nsubN', 'nsubK', 'norb' and 'ndes' in Table 6, together with the final Design-package check on the five orbits B1–B5, are the only evidence excluding 166 candidate tuples, and none of these computations can be reproduced from the manuscript. Please supply the GAP code and the full output/logs, or an independent verification of the enumeration, so that the classification is machine-checkable.","section":"Section 3, Tables 4-6"}],"minor_comments":[{"comment":"The HS parameter set is printed as (176,1100,50,2); the value k=8 is missing and the tuple should be (176,1100,50,8,2).","section":"Abstract"},{"comment":"The symbol H is used both for the subgroup SU3(q):Z2 and for an orbit of length q^3+1 in the paragraph beginning \"Denote by H the ¯H-orbit\"; this overloading is confusing and should be fixed.","section":"Section 2, construction of D0"},{"comment":"The sentence \"Then M = W : C with W a Sylow 2-subgroup of G\" should read \"W a Sylow 2-subgroup of M\", since W has order q^6 and G has a larger 2-part.","section":"Lemma 2.2"},{"comment":"The point stabilizer for M22 is printed as PSU3(4); its order is incompatible with the degree v=22, so this is presumably PSL3(4), and the corresponding row for M22:2 should be checked in the same way.","section":"Table 1, line 2"},{"comment":"Several statements say \"where q+1 a Fermat prime\" or \"where q−1 is a Mersenne prime\"; these are missing verbs or articles and should be copy-edited.","section":"Throughout"}],"recommendation":"major_revision","confidential_remarks":null},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nThe paper does two things: it closes the remaining exceptional-socle gap for flag-transitive 2-designs with λ prime, and it builds a genuinely new infinite family of 2-designs from G2(q). That family is the part worth remembering. The sporadic-socle classification is a large computational census that recovers the three known designs and rules out all other parameter sets.\n\nThe new construction is the strongest section. The coset geometry Cos(G2(q), SU3(q):Z2, [q^6]:Z_{q-1}) has parameters computed from orbit data, not fitted, and the uniqueness proof in Lemmas 2.2--2.4 is coherent. The counting in Lemma 2.3 uses standard orbit-size results from Higman--McLaughlin and Liebeck--Praeger--Saxl, and the arithmetic forces each k_j and r_j cleanly. The repair of [39, Lemma 3.8] is believable: the q even, t=1, ε=− case was indeed underargued, and the new family arises precisely there. I do not see a circular step or a free parameter.\n\nThe soft spots are real but not fatal. First, the sporadic part rests on a GAP 4.12.2 computation using AtlasRep and Design, but no code, logs, or seeds are included. Tables 4--6 record the filters and the 166 surviving candidates, yet the decisive \"ndes\" verdicts on B1--B5 are black boxes. If a subgroup was missed or an AtlasRep label misread, Theorem 1.2 would be wrong. This is a reproducibility gap, not a demonstrated error, but for a theorem whose completeness depends on computation, referees should expect the code. Second, the field-automorphism normalization at the end of the proof of Theorem 1.1 is terse: the claim that φ can be assumed to normalize H0 and Q0 is asserted more than proved. It is probably standard given the uniqueness of conjugacy classes, but it deserves a sentence or two more.\n\nSelf-citation is not a problem here; the cited earlier papers are published and the G2(q) argument depends on them, but it also goes beyond them. The paper is serious, the new design is a real contribution, and the classification claim is plausible. It deserves a serious referee, and the referee should ask for the GAP code and logs, or at least a more detailed reproducibility appendix, before the sporadic part is taken as established.","headline":"The new G2(q) family and the gap repair are the real content; the sporadic completeness proof is a serious computation whose code should be shipped before it becomes fully checkable.","tokens_in":25092,"tokens_out":1953,"would_cite":true,"duration_ms":21643,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["05B05","05B25","20B25","20D08"],"pacs":[],"model":"deepseek-v4-flash","headline":"For exceptional and sporadic socles, flag-transitive point-primitive 2-designs with prime λ reduce to two infinite families and three isolated designs.","keywords":["2-design","flag-transitive","point-primitive","prime λ","exceptional groups of Lie type","sporadic simple groups","Suzuki-Tits ovoid","coset geometry"],"falsifier":"Recompute the sporadic sieve from scratch: for every maximal subgroup $H$ of each almost simple group with sporadic socle, generate all candidate parameters from $r(k-1)=\\lambda(v-1)$, $rv=bk$, and the index conditions, then for the survivors check whether a subgroup $K$ of the required index exists inside a maximal subgroup $N$ and whether its orbits produce the required block set; any flag-transitive point-primitive design outside Table 1 would refute Theorem 1.2. Separately, for Theorem 1.1, compute the coset geometry $\\mathrm{Cos}(G_2(4),\\mathrm{SU}_3(4)\\!:\\!\\mathbb{Z}_2,[4^6]\\!:\\!\\mathbb{Z}_3)$ and check directly whether each pair of points lies in exactly $\\lambda=5$ blocks; a different value would refute Lemma 2.3.","tokens_in":24166,"feed_emoji":"📐","tokens_out":9290,"duration_ms":81787,"temperature":0.7,"pith_summary":"The paper tries to finish the classification of $2$-$ (v,k,\\lambda)$ designs with $\\lambda$ prime when the automorphism group is almost simple, point-primitive, and flag-transitive, and its socle is a finite exceptional or sporadic simple group. It claims that for exceptional socles only two infinite families occur: the Suzuki–Tits ovoid design, with $\\lambda=q-1$ where $q-1$ is a Mersenne prime, and a new design built from the coset geometry $\\mathrm{Cos}(G_2(q),\\mathrm{SU}_3(q)\\!:\\!\\mathbb{Z}_2,[q^6]\\!:\\!\\mathbb{Z}_{q-1})$, with $\\lambda=q+1$ where $q+1$ is a Fermat prime. For sporadic socles it claims that exactly three designs exist, with parameter sets $(12,22,11,6,5)$, $(22,77,21,6,5)$, and $(176,1100,50,8,2)$. If correct, the result completes one major branch of the reduction of flag-transitive designs with prime $\\lambda$ to almost simple and affine groups.","feed_headline":"Prime λ narrows flag-transitive designs to two families","feed_subtitle":"Exceptional socles force the Suzuki–Tits ovoid or a new G2(q) geometry; sporadic socles allow just three designs.","key_machinery":"The mechanism is the coset geometry construction. Given $T=G_2(q)$, $H=\\mathrm{SU}_3(q)\\!:\\!\\mathbb{Z}_2$, and $K=[q^6]\\!:\\!\\mathbb{Z}_{q-1}$, the incidence structure whose points are cosets of $H$ and whose blocks are cosets of $K$, with incidence defined by nonempty intersection, is shown to be a $2$-design. The proof that $\\lambda=q+1$ hinges on the action of $H$ on the coset space: the $H$-orbits have sizes $q^2(q^3+1)$ for all but one orbit, and the last has size $(q^2-1)(q^3+1)$; solving the resulting $1$-design equations forces each block through a point to meet those orbits in sizes $q^2$ and $q^2-1$, respectively, giving exactly $q+1$ blocks through any two points. The uniqueness of $K$ comes from a fixed incident point-line pair in $\\mathrm{PG}_5(q)$ and the structure of the parabolic stabilizer $T_{\\alpha,\\ell}=R\\!:\\!(\\mathbb{Z}(L)\\times F_{q(q-1)})$.","core_discovery":"The central claim is a structural dichotomy, with a new object on one side. For $T\\ne G_2(q)$, earlier work had already handled the exceptional socles, and the paper keeps that conclusion: the only design is the Suzuki–Tits ovoid design in $\\mathrm{PG}_3(q)$, with $T={}^2B_2(q)$, $q=2^{2a+1}\\ge8$ and $\\lambda=q-1$ a Mersenne prime. For $T=G_2(q)$ with $q\\ge4$ even, the paper identifies a gap in that earlier argument and closes it: the only possible design is the coset geometry $\\mathrm{Cos}(T,H,K)$ with $H=\\mathrm{SU}_3(q)\\!:\\!\\mathbb{Z}_2$ and $K=[q^6]\\!:\\!\\mathbb{Z}_{q-1}$, with parameters $(q^3(q^3-1)/2,(q+1)(q^6-1),(q+1)(q^3+1),q^3/2,q+1)$, and $\\lambda=q+1$ must be a Fermat prime. Lemmas 2.3 and 2.4 show this geometry is indeed a $2$-design with $T$ acting flag-transitively and that any design with those parameters is isomorphic to it. For sporadic socles the claim is a finite list: the three designs in Table 1, no more.","pith_inferences":["Editorial inference: the Mersenne condition on $q-1$ and the Fermat condition on $q+1$ suggest that the eventual full classification for prime $\\lambda$ may be cut by elementary prime-number obstructions; in particular, if the list of Fermat primes is finite, the $G_2(q)$ family is finite even though the group-theoretic construction works for all even $q$.","Editorial inference: the construction via a Hermitian unital inside a Desarguesian line spread of $\\mathrm{PG}_5(q)$ is geometric enough that one could try to build the same coset geometry for other Lie-type groups containing $\\mathrm{SU}_3(q)\\!:\\!\\mathbb{Z}_2$ as a maximal subgroup; success would extend the family beyond $G_2(q)$.","Editorial inference: a fully independent recomputation of the sporadic sieve, starting from the 124 candidate parameter sets and the 166 surviving tuples, would turn the finite classification from a recorded computation into a reproducible one; the paper does not include the scripts or logs."],"forward_implications":["If Theorem 1.1 is correct, the exceptional-socle case is closed: any such design is isomorphic to the Suzuki–Tits ovoid design or to the new $\\mathrm{Cos}(G_2(q),\\mathrm{SU}_3(q)\\!:\\!\\mathbb{Z}_2,[q^6]\\!:\\!\\mathbb{Z}_{q-1})$ geometry, with $\\lambda$ a Mersenne or Fermat prime respectively.","The new $G_2(q)$ design exists for every even $q\\ge4$, and whenever $q+1$ is a Fermat prime the full almost simple group $G=T\\!:\\!\\langle\\varphi\\rangle$ also acts flag-transitively on it, as shown at the end of the proof of Theorem 1.1.","If Theorem 1.2 is correct, the sporadic-socle classification has exactly three members; no symmetric design with sporadic socle and prime $\\lambda$ occurs, and the nonsymmetric examples are precisely the three rows of Table 1.","A corollary drawn in the paper, using the companion result for point-imprimitive designs, is that for these socles no flag-transitive point-imprimitive examples with prime $\\lambda$ exist, so point-primitivity is automatic whenever the socle is exceptional or sporadic."],"supporting_citations":[{"why":"Supplies the prior classification for exceptional socles other than $G_2(q)$ and the lemma whose $G_2(q)$ case the paper revisits and corrects.","marker":"[39]"},{"why":"Gives the coset-geometry construction and the parameter formulas for $\\mathrm{Cos}(T,H,K)$ used to establish the new design.","marker":"[24]"},{"why":"Provides the uniqueness and structure of maximal subgroups $\\mathrm{SU}_3(q)\\!:\\!\\mathbb{Z}_2$ and related subgroups used to identify $H$ and $K$.","marker":"[13]"},{"why":"Supplies the structure of the stabilizer $T_{\\alpha,\\ell}$ and the fixed incident point-line pair used to prove uniqueness of $K$.","marker":"[15]"},{"why":"Gives the explicit construction of the Suzuki–Tits ovoid design used for part (a) of Theorem 1.1.","marker":"[3]"},{"why":"Classifies the sporadic socle cases with $\\gcd(r,\\lambda)=1$, yielding the first two rows of Table 1.","marker":"[37]"},{"why":"Provides the design with parameters $(176,1100,50,8,2)$ for the socle HS listed in Table 1.","marker":"[28]"},{"why":"Supplies the orbit-size proposition used to compute the block sizes and force $\\lambda=q+1$ in the new design.","marker":"[30]"}],"fun_headline_variants":["New G2(q) geometry joins flag-transitive design list","Prime λ leaves two exceptional families and three sporadic designs","Exceptional socles give two families, sporadic give three designs","Closing a gap: G2(q) yields a new flag-transitive design","Sporadic socles allow only three flag-transitive designs"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that the finite computation in Section 3—listing maximal subgroups, candidate parameters, subdegrees, subgroups of the required index, and orbit tests—was done correctly and completely; a single missed subgroup or faulty orbit check could admit extra sporadic designs.","fun_headline_variants_meta":{"raw":{"variants":["New G2(q) geometry joins flag-transitive design list","Prime λ leaves two exceptional families and three sporadic designs","Exceptional socles give two families, sporadic give three designs","Closing a gap: G2(q) yields a new flag-transitive design","Sporadic socles allow only three flag-transitive designs"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.002418,"raw_usage":{"total_tokens":9399,"prompt_tokens":1148,"completion_tokens":8251,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":764,"completion_tokens_details":{"reasoning_tokens":8164}},"tokens_in":764,"tokens_out":8251,"duration_ms":57924,"temperature":1.0,"reasoning_tokens":8164,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-15T23:17:08.061795+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Recompute the sporadic sieve from scratch: for every maximal subgroup $H$ of each almost simple group with sporadic socle, generate all candidate parameters from $r(k-1)=\\lambda(v-1)$, $rv=bk$, and the index conditions, then for the survivors check whether a subgroup $K$ of the required index exists inside a maximal subgroup $N$ and whether its orbits produce the required block set; any flag-transitive point-primitive design outside Table 1 would refute Theorem 1.2. Separately, for Theorem 1.1, compute the coset geometry $\\mathrm{Cos}(G_2(4),\\mathrm{SU}_3(4)\\!:\\!\\mathbb{Z}_2,[4^6]\\!:\\!\\mathbb{Z}_3)$ and check directly whether each pair of points lies in exactly $\\lambda=5$ blocks; a different value would refute Lemma 2.3.","supporting_citations":[{"cited_title":"Zhang, J","cited_arxiv_id":null,"evidence_quote":"Supplies the prior classification for exceptional socles other than $G_2(q)$ and the lemma whose $G_2(q)$ case the paper revisits and corrects."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Gives the coset-geometry construction and the parameter formulas for $\\mathrm{Cos}(T,H,K)$ used to establish the new design."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the structure of the stabilizer $T_{\\alpha,\\ell}$ and the fixed incident point-line pair used to prove uniqueness of $K$."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Gives the explicit construction of the Suzuki–Tits ovoid design used for part (a) of Theorem 1.1."},{"cited_title":"Liang, S","cited_arxiv_id":null,"evidence_quote":"Provides the design with parameters $(176,1100,50,8,2)$ for the socle HS listed in Table 1."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Supplies the orbit-size proposition used to compute the block sizes and force $\\lambda=q+1$ in the new design."}],"review_version":1}