{"id":"64137abe-ed0a-4d47-aa02-6e9cf12c157d","arxiv_id":"1908.01649","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":7.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"A two-generated finite group has a planar generating graph if and only if it is one of C2, C3, C4, C5, C6, C2×C2, D3, D4, Q8, C4×C2, or D6.","lead":"This paper classifies all two-generated finite groups whose generating graph, the picture that joins two elements when they generate the whole group, is planar: only eleven groups qualify, the smallest cyclic groups plus a few dihedral and quaternion exceptions. The proof uses the probability that two random elements generate the group and a chief-series factorization to force every planar candidate into the short list.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Section 3's final case falsely asserts α_j≥2 for all j>1; Lemma 4(2) permits α=3/2 for a complemented order-2 chief factor below an odd chief factor, and C2×A4 realizes this, so the t≤2 step is not justified.","rationale":"The paper's central claim is a classification of finite 2-generated groups with planar generating graph. The necessity proof reduces planarity to the bound ∏α_i < 6 and then splits into cases. The most load-bearing issue is in the final split, where the assertion 'α_j ≥ 2 if j > 1' is false. Lemma 4 includes a case α = 3/2 for an order-2 chief factor with a complement and a quotient with no order-2 epimorphic image; the group C2×A4 realizes this configuration, so the proof's step forcing t ≤ 2 is not valid. This is a genuine gap in the printed derivation, but it does not by itself exhibit a planar counterexample: the same witness has ∏α_i = 12, and the product constraint seems to leave only repairable cases. The reader's additional Dic3 concern is not supported, since Dic3 has quotient C4 rather than C2×C2, so it does not threaten the order-12 step. Overall, the verdict remains conditional: the classification is likely correct, but the proof needs a careful repair of the final case analysis.","tokens_in":4823,"tokens_out":41153,"duration_ms":408591,"concrete_test":"Re-derive the final case of Section 3 using Lemma 4's exact trichotomy for order-2 chief factors (α = 1, α = 3/2, or α ≥ 2) instead of the blanket 'α_j ≥ 2'. Show that the product bound ∏α_i < 6 together with the absence of C2×C2 quotients forces either t ≤ 2 or a group excluded by minimality and complement constraints; if the only survivors are C3×C3, D3, and A4, then the false α_j line is a repairable proof defect and the classification stands.","verdict_should_be":"UNCHANGED","load_bearing_attack":"In the last case of Section 3, the proof claims 'α_j ≥ 2 if j > 1' and uses this to force t ≤ 2. This is not a consequence of Lemma 4. Lemma 4(2) explicitly gives α = 3/2 when N is an order-2 chief factor, N has a complement, and G/N has no epimorphic image of order 2. Such a factor can occur below an odd chief factor without forcing a C2×C2 quotient. A concrete witness is G = C2×A4, which is 2-generated, soluble, and has no C2×C2 quotient. A chief series for G has factors C3, V4, C2; the bottom C2 has the single complement A4, and G/N = A4 has no C2 quotient, so Lemma 4(2) gives α = 3/2, not α ≥ 2. Thus the derivation 'By (3.2), we must have t≤2' is invalid, and the printed necessity proof omits the possibility of t≥3 groups in this case. The reader's companion concern about Dic3 does not land: Dic3 has quotient C4, not C2×C2, so it is not a counterexample to the order-12 noncentral-C3 step. But the α_j gap is real: as written, the classification is not fully derived, even though no planar counterexample is presently known.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper classifies the finite 2-generated groups whose generating graph Γ(G) is planar. The main theorem asserts that Γ(G) is planar exactly for the eleven groups C2, C3, C4, C5, C6, C2×C2, D3, D4, Q8, C4×C2, and D6. The proof combines the planar edge bound |E| ≤ 3|V|−6 with probabilistic generation: writing P_G(2) as a product over chief factors, the author defines α_i = |N_{i-1}/N_i| P_{G/N_i, N_{i-1}/N_i}(2) and derives the bound ∏ α_i < 6 from planarity. Lemmas 2–4 give lower bounds on α for nonabelian and abelian chief factors, leading to solubility of any planar-generating 2-generated group. The remaining case analysis handles cyclic groups of order at least 7, groups admitting a C2×C2 quotient, and groups without one; the 'if' direction is established by explicit planar drawings for the listed groups.","tokens_in":5113,"tokens_out":17853,"duration_ms":183052,"significance":"If the proof gap identified below is repaired, the theorem gives a complete, elegant classification of a natural graph invariant of finite groups. The strategy of passing from planarity to the ratio e(G)/|G| and then to chief-series factors is attractive and likely to be reusable. The paper explicitly exhibits planar drawings for all eleven groups, and the main impossibility argument is a clean application of Gaschütz's and Detomi–Lucchini's probabilistic generation results. Those results concern generation probabilities, not planarity, so the classification is not circularly dependent on the conclusion. The principal weakness is a nontrivial missing case in the final paragraph of Section 3; the theorem may well be true, but the printed derivation is incomplete.","major_comments":[{"comment":"The assertion \"α_j ≥ 2 if j > 1\" in the final case (where C2×C2 is not an epimorphic image of G) is not a consequence of Lemma 4. Lemma 4(2) explicitly allows α = 3/2 for an order-2 chief factor N when N has a complement in G/N_j and G/N_j has no epimorphic image of order 2. Such a factor can occur below an odd chief factor without giving a C2×C2 quotient of G. A concrete witness is G = C2×A4: it is 2-generated, has no C2×C2 quotient, and a chief series with factors C3, V4, C2; for the bottom C2 factor, the complement A4 exists and A4 has no C2 quotient, so Lemma 4(2) gives α = 3/2. Consequently the deduction \"By (3.2), we must have t ≤ 2\" is invalid, and the printed proof does not exclude the possibility of t ≥ 3 in this case.","section":"Section 3, final paragraph"},{"comment":"Related to the previous comment, the conclusion \"By Lemma 4, |N1| ≤ 4\" and the subsequent list of remaining possibilities (C3×C3, Alt(4), D3) depend on the invalid t ≤ 2 step. A group such as C2×A4 has a chief series of length 3 and no C2×C2 quotient, so it is not covered by the case analysis as written. The author must either prove that α_j ≥ 2 for j > 1 under an additional hypothesis that holds in this case, or supply a different argument that handles lower order-2 chief factors with a complement and no C2 quotient.","section":"Section 3, final paragraph"}],"minor_comments":[{"comment":"In the sentence \"let e(G) be the number of vertices of the generating graph\", the word \"vertices\" should be \"edges\"; the subsequent formula e(G) = |G|^2 P_G(2)/2 is for edges.","section":"Section 3, opening paragraph"},{"comment":"The expression \"α3 ≥ 2 if j > 2\" contains an index typo; it should read \"α_j ≥ 2 if j > 2\".","section":"Section 3, C2×C2 quotient subcase"},{"comment":"The claim that 2-generation of G forces H/K ≅ 1 or C2 is not immediate; H/K is an elementary abelian 2-group generated by at most two elements, so in principle H/K could be V4. The missing argument is that V4 would give c = 4 and hence P_{G,N}(2) = 0, contradicting the assumption that G is 2-generated.","section":"Lemma 4, proof"},{"comment":"The notation g^n for conjugation is used without explicit definition; it would help to state that g^n = n^{-1} g n, although the displayed equation with commutators makes the intended meaning clear.","section":"Section 2, Lemma 2"}],"recommendation":"major_revision","confidential_remarks":"The paper is by a senior researcher and cites two of the author's earlier results for the probabilistic bounds; I see no circularity concern, as those results are general generation theorems. The Section 3 gap is genuine and load-bearing, but it appears fixable: the author needs to handle groups with a lower order-2 chief factor that is complemented and has no C2 quotient, rather than asserting α_j ≥ 2 for all lower factors. No planar counterexample is currently known, so I do not recommend rejection."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Short version: this is a good paper with a correct-looking classification, but the printed proof is missing a case. The main theorem—planar generating graph iff one of eleven small groups—is exactly the kind of result the subfield wants, and the alpha(G,N) ratio trick is a genuinely useful idea. The positive direction is fine: the explicit drawings for the listed groups are there, and the cyclic and 2-group cases are handled correctly.\n\nThe problem is in the last paragraph of Section 3. After excluding C2×C2 quotients, the author asserts α_j ≥ 2 for every j>1. That is not what Lemma 4 gives. For an order-2 chief factor N that has a complement and whose quotient G/N has no epimorphic image of order 2, Lemma 4(2) gives α=3/2. The proof does not rule this out for lower factors in the final case. A concrete witness is G=C2×A4, which has no C2×C2 quotient. A chief series can be chosen with factors C3, C2, V4; the middle C2 is complemented in G/V4 and the quotient by it is C3, so α=3/2 for that factor. Thus the chain 'α_j≥2 for j>1, so (3.2) forces t≤2' fails. The rest of the case analysis then misses groups like C2×A4. The product bound still gives |G|P_G(2)=13 for this witness, so the theorem is not in immediate danger, but the printed derivation doesn't get you there.\n\nOne side note: the reader's worry about Dic3 does not survive contact with the paper. Dic3 has a quotient C4, not C2×C2, so it does not fit the order-12 noncentral-C3 subcase. That concern should be dropped.\n\nBottom line: the classification is likely correct, the probabilistic machinery is sound, and the gap is local. But it is load-bearing: without fixing the α_j≥2 claim, the necessity proof is incomplete. This deserves a proper referee, and the author should be asked to patch Section 3.","headline":"A clean classification, but the necessity proof has a real gap: the final case wrongly asserts α_j≥2 for lower chief factors, and C2×A4 shows the claim is false.","tokens_in":5606,"tokens_out":17805,"would_cite":false,"duration_ms":162083,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["20D60","05C25","20P05"],"pacs":[],"model":"deepseek-v4-flash","headline":"A finite 2-generated group has a planar generating graph exactly when it is one of eleven groups.","keywords":["generating graph","planar graph","finite 2-generated groups","chief series","probabilistic generation","minimal normal subgroups","K5 subdivision","K3,3 subdivision"],"falsifier":"Construct the generating graph of the dicyclic group of order 12, the non-dihedral group with a normal $C_3$ and a $C_2\\times C_2$ quotient, and check whether it is planar. The proof excludes this group only through the assertion that such a group is $D_6$, so a planar drawing of its graph would refute the classification, while a $K_5$ or $K_{3,3}$ subdivision would show that the final-case argument needs to identify that case explicitly.","tokens_in":4607,"feed_emoji":"📐","tokens_out":10823,"duration_ms":96313,"temperature":0.7,"pith_summary":"The paper establishes a complete classification: a finite group that can be generated by two elements has a planar generating graph exactly when it is one of eleven groups. The generating graph puts a vertex at every non-identity element and joins two vertices when those elements together generate the group; planarity means the graph can be drawn without crossing edges. The eleven permitted groups are the cyclic groups $C_2$, $C_3$, $C_4$, $C_5$, $C_6$, the Klein four-group $C_2\\times C_2$, the dihedral groups $D_3$, $D_4$, $D_6$, the quaternion group $Q_8$, and $C_4\\times C_2$. For every other finite 2-generated group the paper claims the generating graph contains a subdivision of $K_5$ or $K_{3,3}$, the two minimal forbidden configurations for planar graphs. The value of such a classification is that a geometric property of a large auxiliary graph becomes a finite checklist on the group's structure.","feed_headline":"Eleven 2-generated groups have planar generating graphs","feed_subtitle":"Planar generating graphs occur only for eleven small groups; all others force a K5 or K3,3 subdivision.","key_machinery":"The workhorse is the ratio $\\alpha(G,N)=\\frac{e(G)/|G|}{e(G/N)/|G/N|}$ for a minimal normal subgroup $N$, equal to $|N|P_{G,N}(2)$, the conditional probability that two random elements generate $G$ given that they generate $G/N$. Along a chief series these factors multiply to $|G|P_G(2)$, twice the expected edge count per vertex. Lemma 4, built on the formula $P_{G,N}(2)=1-c/p^{2a}$, says the factor is $1$, $\\frac32$, or at least $2$, with the small values reserved for order-2 complemented chief factors. Comparing the product of these factors with the planar edge bound $3n-6$ is what shrinks the universe of possible groups to a finite list, after which explicit drawings finish the proof.","core_discovery":"The central claim is the bi-conditional: if $G$ is a finite 2-generated group, then $\\Gamma(G)$ is planar if and only if $G$ belongs to the eleven-element list $C_2, C_3, C_4, C_5, C_6, C_2\\times C_2, D_3, D_4, Q_8, C_4\\times C_2, D_6$. The proof splits into a universal exclusion and an explicit exhibition. The exclusion starts from the planar edge bound $e\\le 3|G|-6$, which forces the product of chief-series ratios $\\alpha_i$ to be less than 6, and shows via the ratio lemmas that this can only happen for groups built from very small chief factors; the surviving candidates are then eliminated one by one, except for the listed groups, whose generating graphs are drawn in the plane. Since the listed graphs are planar by construction, the classification is complete if every borderline case in the final count is genuinely excluded.","pith_inferences":["The same chief-series ratio technique should classify other sparse graph properties of generating graphs, such as bounded genus or bounded treewidth, by substituting the appropriate edge bound for $3n-6$; the paper's structure already delivers the needed product formula.","A computer search over 2-generated groups of small order could test the borderline cases directly, checking whether any group outside the eleven-list has a planar generating graph and thereby sharpening or confirming the final-case analysis.","The dependency on the conditional probability $P_{G,N}(2)$ suggests a probabilistic reading of planarity: only groups whose two-element generation probability decays slowly relative to group order can afford the low edge counts that planarity permits."],"forward_implications":["Any finite 2-generated group outside the eleven-group list has a generating graph that is nonplanar, hence contains a $K_5$ or $K_{3,3}$ subdivision.","For the eleven groups, planarity can be certified by explicit drawings; the non-cyclic order-8 cases $D_4$, $Q_8$, and $C_4\\times C_2$ have isomorphic planar skeletons after isolated vertices are deleted.","The inequality $|G|P_G(2)<6$ becomes a necessary test: any group with two-element generation probability above $6/|G|$ cannot have a planar generating graph.","The proof reduces a global geometric question to finitely many chief-series computations, so the planarity of $\\Gamma(G)$ is decidable for any given finite 2-generated group by checking a short list."],"supporting_citations":[{"why":"Supplies the planar graph edge bound $3n-6$ used to turn planarity into the inequality $|G|P_G(2)<6$.","marker":"[1]"},{"why":"Gives the reduction of the conditional generation probability for soluble quotients used in Lemma 3.","marker":"[4]"},{"why":"Provides the lower bound $\\frac{53}{90}$ for the conditional probability in the unique minimal normal subgroup case.","marker":"[5]"},{"why":"Establishes that the size of the fibre over a generating tuple modulo $N$ is independent of the tuple, so the ratio $\\alpha(G,N)$ is well defined.","marker":"[6]"},{"why":"Gives the formula $P_{G,N}(2)=1-c/p^{2a}$ that drives Lemma 4 and the small-factor classification.","marker":"[7]"}],"fun_headline_variants":["Eleven groups: the full list for planar generating graphs","Planar generating graphs: exactly eleven 2-generated groups","When does a generating graph embed in a plane? Only for 11 groups","A short classification: planar generating graphs for 2-generated groups"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"Two load-bearing assertions carry the final exclusion: that a chief factor of order 2 with no $C_2$ quotient above it contributes a factor at least 2, where the ratio lemma only guarantees $\\frac32$, and that the only order-12 group with a noncentral normal $C_3$ and a $C_2\\times C_2$ quotient is $D_6$. If either admits another case, the classification could be incomplete.","fun_headline_variants_meta":{"raw":{"variants":["Eleven groups: the full list for planar generating graphs","Planar generating graphs: exactly eleven 2-generated groups","When does a generating graph embed in a plane? Only for 11 groups","A short classification: planar generating graphs for 2-generated groups"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000829,"raw_usage":{"total_tokens":3582,"prompt_tokens":863,"completion_tokens":2719,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":479,"completion_tokens_details":{"reasoning_tokens":2647}},"tokens_in":479,"tokens_out":2719,"duration_ms":19136,"temperature":1.0,"reasoning_tokens":2647,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T15:10:31.617113+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Construct the generating graph of the dicyclic group of order 12, the non-dihedral group with a normal $C_3$ and a $C_2\\times C_2$ quotient, and check whether it is planar. The proof excludes this group only through the assertion that such a group is $D_6$, so a planar drawing of its graph would refute the classification, while a $K_5$ or $K_{3,3}$ subdivision would show that the final-case argument needs to identify that case explicitly.","supporting_citations":[{"cited_title":"Aschbacher and R","cited_arxiv_id":null,"evidence_quote":"Supplies the planar graph edge bound $3n-6$ used to turn planarity into the inequality $|G|P_G(2)<6$."},{"cited_title":"Breuer, R","cited_arxiv_id":null,"evidence_quote":"Gives the reduction of the conditional generation probability for soluble quotients used in Lemma 3."},{"cited_title":"Ballester-Bolinches and L","cited_arxiv_id":null,"evidence_quote":"Provides the lower bound $\\frac{53}{90}$ for the conditional probability in the unique minimal normal subgroup case."},{"cited_title":"Breuer, R","cited_arxiv_id":null,"evidence_quote":"Establishes that the size of the fibre over a generating tuple modulo $N$ is independent of the tuple, so the ratio $\\alpha(G,N)$ is well defined."},{"cited_title":"Burness and E","cited_arxiv_id":null,"evidence_quote":"Gives the formula $P_{G,N}(2)=1-c/p^{2a}$ that drives Lemma 4 and the small-factor classification."}],"review_version":1}