{"id":"a8f61fb5-23ae-479b-9b20-ec9d9de91e58","arxiv_id":"2608.10035","paper_version":1,"verdict":"ACCEPT","confidence":"HIGH","novelty_score":6.0,"correctness_risk":"low","formal_verification":"none","parameter_count":0,"one_line_summary":"For a finite group with a normal subgroup, two new relative invariants of element orders are shown to lie on opposite sides of the corresponding invariants of the quotient, with equality cases characterized as equal order pairs.","lead":"This paper introduces two new ways to measure the average order of elements of a finite group that lie outside a normal subgroup, and proves that these measurements obey clean inequalities when compared with the same measurement on the quotient group. It also describes exactly which groups achieve equality.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"No significant objection identified. The main proof is sound; the only external dependency is the cited classification [19, Thm. 1.1], whose hypotheses appear to match the argument.","rationale":"After rechecking the proof, the inequalities are correct: with m=[G:H], the lower bound chain sum_i o(y_iH) leq (psi(G)-psi(H))/|H| gives inequality (1), and the upper bound chain gives inequality (2). In Lemma 2.2, the identity o(x)=o(xH)o(x^n) is valid because n=o(xH) divides o(x), so o(x^n)=|H| and H=langle x^nrangle leq langle xrangle; hence H is contained in every cyclic subgroup generated by an element outside H, i.e., H is a breaking point in the poset of cyclic subgroups. The cited theorem [19] yields the claimed dichotomy, and the converse cases are routine. The dependence on [19] is real, but it is a published theorem with matching hypotheses; absent evidence of a misstatement or a hidden counterexample, it does not undermine the central claim. The reader's ACCEPT verdict with high confidence remains appropriate.","tokens_in":5657,"tokens_out":23755,"duration_ms":227609,"concrete_test":"Verify [19, Theorem 1.1] by reading its statement and proof: confirm it classifies finite groups with a nontrivial H comparable with every cyclic subgroup as cyclic p-groups of order at least p^2 or generalized quaternion 2-groups. If the theorem instead requires comparability with all subgroups, check whether its proof still covers the cyclic-subgroup comparability used in Lemma 2.2. As an independent numerical check, use GAP SmallGroups to enumerate all groups of order at most 64 and all nontrivial proper normal subgroups H, and confirm that the pairs satisfying equality in (2) coincide exactly with the claimed cyclic p-group cases and Q_{2^n} cases.","verdict_should_be":"UNCHANGED","load_bearing_attack":"No significant objection identified. The inequalities in Theorem 1.1 follow from a clean coset-wise min/max argument, and the equality characterizations are locally sound. The only non-self-contained step is Lemma 2.2, which invokes [19, Theorem 1.1] to pass from 'H is comparable with every cyclic subgroup' to the dichotomy cyclic p-group / generalized quaternion 2-group. The hypotheses derived in the proof do imply the needed comparability: from condition (4) and the identity o(x)=o(xH)o(x^n), one obtains o(x^n)=|H|, hence H=langle x^nrangle leq langle xrangle for every x outside H, making H a breaking point in the poset of cyclic subgroups. The cited theorem is a published classification with apparently matching hypotheses; I found no internal inconsistency, misapplication, or counterexample in the small cases checked conceptually.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper introduces two relative invariants for a finite group G and a proper normal subgroup H: o(G,H) = (ψ(G)-ψ(H))/(|G|-|H|) and ψ''(G,H) = (ψ(G)-ψ(H))/(|G|^2-|H|^2). The main result, Theorem 1.1, proves that o(G,H) ≥ o(G/H,1) and ψ''(G,H) ≤ ψ''(G/H,1), with equality characterized: equality in the first inequality holds exactly when (G,H) is an equal order pair with H an isolated subgroup, and equality in the second holds exactly when G is a cyclic p-group with H a proper nontrivial subgroup or G is a generalized quaternion 2-group with H its center. The proof of the inequalities uses a coset decomposition and min/max bounds; the equality cases reduce to lemmas, one of which relies on a published classification. Two corollaries apply the inequalities to metabelian groups and ZM-groups, and an open problem is posed about equal order pairs with isolated subgroups.","tokens_in":5825,"tokens_out":27377,"duration_ms":242883,"significance":"The central inequalities are correct and are proved by a short, self-contained min/max argument that is a natural extension of the standard sum-of-element-orders machinery. The equality characterizations are nontrivial and connect the new invariants to equal order pairs, isolated subgroups, and the known dichotomy of cyclic p-groups and generalized quaternion 2-groups. The paper is therefore a useful contribution to the literature on sums of element orders, even though the applications are modest and the main theorem is local in scope. The proof of the equality case in (2) depends on the author's earlier characterization [19]; I checked the hypotheses and the application appears sound for nontrivial H.","major_comments":[],"minor_comments":[{"comment":"As stated, Lemma 2.2 is false when H=1: for every finite group G, (G,1) is an equal order pair satisfying (4), yet G need not be a cyclic p-group or a generalized quaternion 2-group. Please add the hypothesis that H is non-trivial to the statement (and to the proof, where the appeal to [19] requires a nontrivial breaking point).","section":"Lemma 2.2"},{"comment":"These statements need the hypotheses m>1 and n>1. For m=1, ZM(1,n,r) is cyclic of order n and H=1; if n is a prime power, the claimed strict upper bound in Corollary 1.3 becomes an equality, and the equivalence in Lemma 2.1 fails because d=1 while 'G is a Frobenius group with kernel H' is false for a trivial kernel. The degenerate cases should be excluded or handled separately.","section":"Lemma 2.1 and Corollary 1.3"},{"comment":"The step with the notation '⟨b^x a^y⟩ = ⟨b^x⟩ α(u,v)' and the subsequent expression 'b^x a^y = b^{xz} a^{v(1-r^{xz})}' is very hard to follow and appears to contain a typo in the normal-form computation. Please rewrite this part with explicit normal forms and justify why z=1 follows.","section":"Lemma 2.1, proof of a)⇒b)"},{"comment":"The notation 'H = p^{n-i}G' is nonstandard; it would be clearer to write that H is the unique subgroup of order p^i, or H = ⟨p^{n-i}⟩ in the cyclic group C_{p^n}.","section":"Theorem 1.1 statement"},{"comment":"The phrase 'Let G is a finite metabelian group' should be 'Let G be a finite metabelian group'.","section":"Corollary 1.2"},{"comment":"The sentence about ψ''(G,H) providing 'finer structural control when investigating the boundary cases of Camina pairs and Frobenius groups' is vague; a concrete explanation of why the square normalization distinguishes the equality cases would help orient the reader.","section":"Introduction"}],"recommendation":"minor_revision","confidential_remarks":"The main theorem is sound and the central derivation is correct. The issues that require attention are local: Lemma 2.2 needs the nontriviality of H, and Lemma 2.1/Corollary 1.3 need m,n>1 to avoid degenerate counterexamples. The dependence on [19] is legitimate and the application of that classification is appropriate for nontrivial H. I recommend minor revision rather than major revision because the fixes are small and do not affect the main inequalities or the equality characterization as stated in Theorem 1.1."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"The paper is a short, correct note in the sum-of-element-orders area. The new content is the pair of relative invariants o(G,H) and psi''(G,H), defined on a finite group with a proper normal subgroup, and Theorem 1.1 stating o(G,H) >= o(G/H,1) and psi''(G,H) <= psi''(G/H,1), with equality characterizations. The proof is a clean coset-wise min/max argument on the orders of elements in nontrivial cosets. I verified the steps: the inequalities reduce to comparing the min/max representatives to the quotient, and the equality cases follow by translating the equalities back to conditions on every element outside H. The first equality becomes 'equal order pair with H isolated'; the second becomes 'equal order pair satisfying o(x)=o(xH)|H|', and then Lemma 2.2 gives the dichotomy cyclic p-group / generalized quaternion 2-group. That lemma is the one non-self-contained step: it invokes Tarnauceanu's earlier characterization [19, Thm 1.1] of groups with a subgroup comparable to every cyclic subgroup. The dependency is load-bearing but the hypotheses are satisfied, and I found no misapplication. The classification itself is published peer-reviewed work, so this is a reasonable citation rather than circularity.\n\nThe corollaries for metabelian groups and ZM-groups follow directly. The ZM-group equality condition in Corollary 1.3 uses Lemma 2.1, whose proof is terse but valid: the equivalence between equal order pair and Camina pair there uses the known fact that cyclic subgroups in a ZM-group are conjugate exactly when they have equal orders, and the algebra checks out. The open problem at the end is natural.\n\nSoft spots: this is a note, not a deep theory. The impact is limited to the specialized community working on psi(G) and related invariants. The reliance on [19] and [6] means a referee should double-check those classifications, but on inspection they fit. There are no data or code to evaluate. The paper self-cites the author's previous work heavily, but in context these citations are relevant and not circular.\n\nThe paper deserves a serious referee. I would accept it if the editor's scope includes short technical notes in group theory. It is not a major advance, but the result is correct, cleanly proved, and genuinely new.","headline":"Correct, compact note introducing two relative invariants for sums of element orders; the main inequalities and equality cases hold, with an external classification that checks out.","tokens_in":6347,"tokens_out":3944,"would_cite":true,"duration_ms":34795,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["20D60","20E34"],"pacs":[],"model":"deepseek-v4-flash","headline":"This note introduces two normalised sums of element orders attached to a normal subgroup and proves that one is bounded below and the other above by the corresponding quotient value, with equality fully classified.","keywords":["element orders","finite groups","equal order pairs","isolated subgroups","sum of element orders","average element order","cyclic p-groups","generalized quaternion 2-groups"],"falsifier":"Check all finite groups of order at most 64 by machine: for every proper nontrivial normal subgroup $H$, compare $\\psi''(G,H)$ with $\\psi''(G/H,1)$. Any equality outside the pairs $(C_{p^n}, C_{p^i})$ and $(Q_{2^n}, Z(Q_{2^n}))$ would refute the stated classification of equality cases; separately, a group with a nontrivial subgroup comparable with every cyclic subgroup but not cyclic or quaternion would refute the cited classification used in Lemma 2.2.","tokens_in":5480,"feed_emoji":"🧮","tokens_out":16211,"duration_ms":147143,"temperature":0.7,"pith_summary":"This note studies two numbers attached to a finite group $G$ and a proper normal subgroup $H$: the average order of the elements outside $H$, and a variant that divides the same sum by $|G|^2-|H|^2$. The main theorem proves that the first number cannot fall below the corresponding average in the quotient $G/H$, while the second cannot exceed its quotient analogue. It then characterizes exactly when equality occurs: for the first, precisely when $(G,H)$ is an equal order pair with $H$ an isolated subgroup; for the second, precisely when $G$ is a cyclic $p$-group with any proper nontrivial $H$, or a generalized quaternion $2$-group with $H$ its center. These inequalities yield two-sided bounds on $\\psi(G)$ for group extensions, and the paper works out the case of ZM-groups as an example.","feed_headline":"Two new inequalities pin element-order sums between quotient bounds","feed_subtitle":"Two new averages of element orders are bounded by quotient values; equality cases are classified.","key_machinery":"The argument is carried by the coset decomposition $\\psi(G)=\\psi(H)+\\sum_{i=2}^{m}\\sum_{h\\in H}o(x_i h)$, where $m=[G:H]$ and $x_iH$ runs through the non-trivial cosets of $H$, together with the choice of elements $y_i,z_i$ in each coset attaining the minimum and maximum of $o(x_i h)$. Sandwiching each inner sum between $o(y_i)|H|$ and $o(z_i)|H|$, then estimating $o(y_i)\\ge o(y_iH)$ and $o(z_i)\\le o(z_iH)|H|$, proves both inequalities directly. The equality analysis turns the collapsed sandwich into the pointwise conditions (3) and (4), identifies them as equal-order-pair conditions, and for the second inequality invokes Lemma 2.2: condition (4) forces $\\langle x^n\\rangle=H\\subseteq\\langle x\\rangle$, so $H$ is comparable with every cyclic subgroup, and the cited classification restricts $G$ to cyclic $p$-groups and generalized quaternion $2$-groups.","core_discovery":"The central claim is a pair of sharp inequalities for the normalised sums $o(G,H)=(\\psi(G)-\\psi(H))/(|G|-|H|)$ and $\\psi''(G,H)=(\\psi(G)-\\psi(H))/(|G|^2-|H|^2)$, valid for every finite group $G$ and proper normal subgroup $H$: $o(G,H)\\ge o(G/H,1)$ and $\\psi''(G,H)\\le \\psi''(G/H,1)$. With $H$ non-trivial, equality in the first holds exactly when the elements of each coset $xH$ all have the same order as $x$ and $H$ meets every cyclic subgroup outside itself trivially, i.e. $(G,H)$ is an equal order pair with $H$ isolated. Equality in the second holds exactly when $G\\cong C_{p^n}$ with $H$ an arbitrary proper nontrivial subgroup, or $G\\cong Q_{2^n}$ with $H=Z(G)\\cong C_2$.","pith_inferences":["Editorial extension: the same coset-sandwich should apply to a chain of normal subgroups, giving telescoping products of relative averages and hence bounds on $\\psi(G)$ for iterated extensions; the note treats only a single subgroup $H$.","Editorial extension: condition (4) could be studied without the equal-order-pair assumption; if equality in (2) already forces the two listed families under weaker hypotheses, that would show which part of Lemma 2.2 is really essential.","Editorial extension: the open problem on isolated-subgroup equal order pairs could be approached by tabulating normal subgroups $H$ of small groups such that $H$ is isolated; such a census would indicate how close the equality cases are to Frobenius groups."],"forward_implications":["Under the hypotheses of Corollary 1.2, the theorem gives explicit bounds on $\\psi(G)$: $\\psi(H)+|H|(\\psi(G/H)-1)\\le \\psi(G)\\le \\psi(H)+|H|^2(\\psi(G/H)-1)$.","For $G=ZM(m,n,r)$ and $H=\\langle a\\rangle$, the bounds become $\\psi(C_m)+m(\\psi(C_n)-1)\\le \\psi(G)<\\psi(C_m)+m^2(\\psi(C_n)-1)$, with the lower bound an equality exactly under the arithmetic condition $d=n$ and $m_1\\nmid r^{n_1}-1$ for all proper divisors $m_1,n_1$.","Equality in the first inequality characterizes isolated subgroups through a purely coset-order condition, giving a testable way to recognize them.","Equality in the second inequality isolates cyclic $p$-groups and generalized quaternion $2$-groups as the only extremal cases, i.e. the finite groups with a unique minimal subgroup."],"supporting_citations":[{"why":"Introduces the sum of element orders and the cyclic maximum theorem that motivates the two normalised quantities.","marker":"[1]"},{"why":"Supplies the Camina/Frobenius equivalence used in the ZM-group example to identify equal order pairs.","marker":"[6]"},{"why":"Defines equal order pairs, the class of pairs appearing in both equality characterisations.","marker":"[7]"},{"why":"Provides the classification of groups with a nontrivial subgroup comparable with every cyclic subgroup, the load-bearing step in Lemma 2.2 for the second equality case.","marker":"[19]"},{"why":"Introduces the earlier normalized invariant psi''(G) = psi(G)/|G|^2 that the new relative function extends.","marker":"[20]"}],"fun_headline_variants":["Equal-order pairs locked by new element-sum bounds","Sharp bounds tie element orders to quotient structure","New element-sum inequalities classify equality cases","Element-order sums squeezed between quotient bounds","Equality in element-order bounds pinned to exceptional pairs"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is the cited classification, not proved in this note, that a finite group with a nontrivial subgroup comparable with every cyclic subgroup under inclusion must be a cyclic $p$-group or a generalized quaternion $2$-group; if that theorem is wrong or its hypotheses are not met, the only-if direction of the equality characterization for $\\psi''$ collapses.","fun_headline_variants_meta":{"raw":{"variants":["Equal-order pairs locked by new element-sum bounds","Sharp bounds tie element orders to quotient structure","New element-sum inequalities classify equality cases","Element-order sums squeezed between quotient bounds","Equality in element-order bounds pinned to exceptional pairs"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001021,"raw_usage":{"total_tokens":4242,"prompt_tokens":812,"completion_tokens":3430,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":428,"completion_tokens_details":{"reasoning_tokens":3363}},"tokens_in":428,"tokens_out":3430,"duration_ms":23522,"temperature":1.0,"reasoning_tokens":3363,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T04:14:21.788009+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Check all finite groups of order at most 64 by machine: for every proper nontrivial normal subgroup $H$, compare $\\psi''(G,H)$ with $\\psi''(G/H,1)$. Any equality outside the pairs $(C_{p^n}, C_{p^i})$ and $(Q_{2^n}, Z(Q_{2^n}))$ would refute the stated classification of equality cases; separately, a group with a nontrivial subgroup comparable with every cyclic subgroup but not cyclic or quaternion would refute the cited classification used in Lemma 2.2.","supporting_citations":[{"cited_title":"Amiri, S.M","cited_arxiv_id":null,"evidence_quote":"Introduces the sum of element orders and the cyclic maximum theorem that motivates the two normalised quantities."},{"cited_title":"Camina,Some conditions that almost characterize Frobenius groups, Israel J","cited_arxiv_id":null,"evidence_quote":"Supplies the Camina/Frobenius equivalence used in the ZM-group example to identify equal order pairs."},{"cited_title":"Camina, R.D","cited_arxiv_id":null,"evidence_quote":"Defines equal order pairs, the class of pairs appearing in both equality characterisations."},{"cited_title":"T˘ arn˘ auceanu,A characterization of generalized quaternion2-groups, C.R","cited_arxiv_id":null,"evidence_quote":"Provides the classification of groups with a nontrivial subgroup comparable with every cyclic subgroup, the load-bearing step in Lemma 2.2 for the second equality case."},{"cited_title":"T˘ arn˘ auceanu,Detecting structural properties of finite groups by the sum of element orders, Israel J","cited_arxiv_id":null,"evidence_quote":"Introduces the earlier normalized invariant psi''(G) = psi(G)/|G|^2 that the new relative function extends."}],"review_version":1}