{"id":"95848964-bb4d-4108-9ac0-35f2ddfa4579","arxiv_id":"2607.29324","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Every regular Borel subalgebra of Lie(Aut(A^2)) is either the metabelian Lie(∆,ℓ(1,1)) or a derived-length-3 algebra Lie(t2,ℓ(a,b),(a,b)) with a≠b, classified up to graded isomorphism by n=gcd(a,b).","lead":"This paper classifies all \"regular\" Borel subalgebras — maximal solvable subalgebras generated by homogeneous derivations — of the Lie algebra of polynomial vector fields on the affine plane. It proves each has derived length 2 or 3, identifies the unique metabelian one, and describes the isomorphism classes.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Central classification depends on Lemma 5.10, whose only-if proof is a long case analysis with ambiguous line notation; a missed configuration there would invalidate Theorem 6.1.","rationale":"The reader identifies Lemma 5.10 as the weakest assumption, and I agree: Theorem 6.1's structural dichotomy between derived length 2 and 3 is obtained by feeding Lemma 5.10 into B^(1). The proof of Lemma 5.10 is a multi-case argument whose correctness is hard to certify because the paper's notation does not reliably distinguish three different line constructs. I did not find a concrete counterexample, and the constructive parts of the paper (e.g., Lemma 5.6's explicit bracket chains) give independent support, so I would not reject the central claim. However, the verification gap is real and is exactly what should be checked before the classification is accepted as sound. The reader's additional objection to Lemma 6.5(a) concerns the isomorphism classification, not Theorem 6.1, so it does not change the primary load-bearing risk. Thus the appropriate verdict remains conditional, i.e., unchanged from the reader's CONDITIONAL.","tokens_in":25588,"tokens_out":34563,"duration_ms":277991,"concrete_test":"Rewrite Lemma 5.10's proof with three explicit symbols—say ℓ^0, ℓ^+, ℓ^∥—for the origin line, the line through (−1,−1) and (a,b), and the line through (−1,−1) parallel to ℓ(a,b). Then verify, without relying on typographic ambiguity, the four-case dichotomy for L1 = Lie({(c,d),(e,f)}) and confirm that each excluded case forces a non-solvable subalgebra. As a computational cross-check, implement the bracket formula (2) and enumerate all S ⊆ M with |S| ≤ 4 and coordinates in [−1,8]^2; for each S, compute the derived series of the truncated graded Lie algebra and compare termination with Lemma 5.10's criterion. Any mismatch would give a concrete counterexample; if none appears, the lemma is supported.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Theorem 6.1(3) reduces every derived-length-3 regular Borel subalgebra to the form Lie(t2, ℓ(a,b), (a,b)) by applying Lemma 5.10 to B^(1): the lemma asserts that a non-Abelian solvable Lie(S) must have all elements of S except one lying on a single line through (−1,−1) parallel to ℓ(a,b). The proof of the only-if direction is the load-bearing case analysis: it chooses a non-Abelian solvable pair (a,b),(c,d), then for any new (e,f) examines L1 = Lie({(c,d),(e,f)}) and rules out all but one possibility. The text renders the three distinct line types (origin line, line through (−1,−1) and (a,b), and line through (−1,−1) parallel to ℓ(a,b)) with the same symbol ℓ in several places. In particular, the surviving case is written as (e,f) ∈ ℓ(c,d) = ℓ(a,b); this equality is only correct if ℓ(c,d) is the line through (−1,−1) and (c,d), which coincides with the required parallel line because (c,d) was chosen on it. If the notation is read in the more natural typeset way, the equality is false. The three excluded cases (exceptional pair, (e,f) on the parallel line through (c,d), and (c,d) on the parallel line through (e,f)) each depend on Lemma 5.6 or Lemma 5.9 to produce a non-solvable subalgebra. The paper does not pin down these cases with unambiguous symbols, and the case analysis is long enough that a reader cannot easily verify that no exceptional configuration was overlooked. Since Lemma 5.10 is used directly to force the shape of B^(1), a gap here is a direct threat to the main classification, not merely to the isomorphism section.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper classifies all regular Borel subalgebras of Lie(Aut(A^2)), i.e., maximal solvable subalgebras generated by homogeneous derivations with respect to the standard Z^2-grading. The main result (Theorem 6.1) states that every such subalgebra has derived length 2 or 3; the only metabelian one is Lie(Δ,ℓ(1,1)), and the length-3 ones are exactly Lie(t2,ℓ(a,b),(a,b)) for (a,b)∈Z_{\\ge0}^2∪{(0,−1),(−1,0)} with a≠b. The proof reduces the problem to a classification of regular Abelian and regular solvable subalgebras (Sections 4–5), then uses maximality to identify the Borel subalgebras. The paper also gives explicit presentations of all regular Borel subalgebras in a common format (Lemma 6.4) and a partial isomorphism classification (Proposition 6.6).","tokens_in":26032,"tokens_out":22754,"duration_ms":196668,"significance":"If correct, this is a substantial contribution to the structure theory of the Lie algebra of polynomial vector fields on the affine plane. It goes well beyond the known triangular Borel subalgebras, gives a complete list of all homogeneous-generated Borel subalgebras, and identifies an explicit example not coming from a Borel subgroup. The proof strategy is sound in outline: Lemma 3.7 reduces regular Borel subalgebras to maximal regular solvable subalgebras, and the later structure lemmas are natural. The paper is generally well organized, and the external results it uses are clearly cited. However, the verification burden falls on two case analyses whose notation is not sufficiently controlled, and one of the isomorphism proof's generating-set assertions is concretely false. These issues need to be fixed before the main claims can be fully trusted.","major_comments":[{"comment":"The only-if direction of Lemma 5.10 is the load-bearing step for the main classification. In Case II the proof reduces L1=Lie({(c,d),(e,f)}) to four alternatives and concludes that the surviving case is (e,f)∈ℓ(c,d)=ℓ(a,b). This equality is correct only if ℓ(c,d) denotes the line through (−1,−1) and (c,d) and (c,d) was chosen on the parallel line; under the natural reading of the typesetting, the same symbol is also used for the line through (−1,−1) parallel to ℓ(c,d), in which case the equality is false. The same ambiguity affects the applications of Lemma 5.9 and Lemma 5.6 in Cases (b) and (c). Since Theorem 6.1(3) uses Lemma 5.10 to force B^(1)⊆Lie(ℓ(a,b),(a,b)), this is a direct threat to the classification. Distinct symbols must be introduced for the four line types and the case analysis re-verified in detail.","section":"§5.2, Lemma 5.10 and Theorem 6.1(3)"},{"comment":"The generating-set assertions are false as stated. For n=−1, the set {Y1,Y2,Xi | i≥r} with r≥1 does not generate X0: every defining bracket involving X0 either requires X0 on the left or would produce X0 only from [X0,X1], which again requires X0. For n≥1, the set {Y1,Y2,X0,…,X_{n−1}} does not generate X_n, because [X0,Xi]=γ_i X_{i+n} only raises indices and [Yj,Xi] only scales. Hence the proof that G(−1, ·) and G(n, ·) are not graded isomorphic is unsupported; this is used in Proposition 6.6(b). The conclusion is probably salvageable by a correct finite-generation argument, but the text must be corrected.","section":"§6.2, Lemma 6.5(a)"}],"minor_comments":[{"comment":"In the displayed definition of the maps, φ(Y1) is written twice; the second occurrence should be φ(Y2).","section":"§6.2, Lemma 6.5(b)"},{"comment":"The four line types ℓ(a,b), ℓ(a,b), ℓ(a,b), and eℓ(a,b) are not visually distinguished in the manuscript. Even if the intended reading is clear to the author, the typesetting must be changed (e.g., subscripts, superscripts, or different accents) so that a reader can follow the case analyses in Sections 5 and 6.","section":"§2, Notation 5–8"},{"comment":"The exclusion of (a,b)∈{(m,−1),(−1,m) | m≥1} is compressed. Please spell out that ℓ(a,b)∩M=∅ forces S\\{(a,b)}=∅, contradicting the assumption that B^(1) is non-abelian.","section":"§6.1, Theorem 6.1(3) proof"},{"comment":"The case (a,b)∈{(0,−1),(−1,0)} is written in a compressed form. Since ℓ(0,−1) and ℓ(−1,0) are different lines, the two inclusions into A should be displayed separately to avoid confusion.","section":"§6.2, Proposition 6.6(a)"}],"recommendation":"major_revision","confidential_remarks":"I did not find a counterexample to the main theorem, and the overall approach seems sound. The decisive issues are the unverifiable notation in the Lemma 5.10 case analysis and the incorrect generating-set argument in Lemma 6.5(a). If these are fixed, the paper could be acceptable; I would not reject on the current evidence."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Dear colleague,\n\nThe main theorem here is real and worth engaging with. Theorem 6.1 gives the first classification of regular Borel subalgebras of Lie(Aut(A2)): derived length 2 or 3, a unique metabelian one, and explicit generators for the derived-length-3 cases. The reduction of regular Borel subalgebras to maximal regular solvable subalgebras (Lemma 3.7) and the solvability criteria (Proposition 5.11) are substantial and appear correct. The main proof is a long case analysis, but I did not find a hole in it.\n\nThe soft spots are concentrated in the isomorphism section. Lemma 6.5(a) contains a concrete false claim: for n=-1, the set {Y1,Y2,Xi | i≥r} does not generate G(-1,...) for r≥1, because without X0 the commuting X_i (i≥1) can never produce X0. The companion assertion that {Y1,Y2,X0,...,X_{n-1}} is the smallest generating set for n≥1 is also asserted without proof and needs hypotheses on the γ_i. Proposition 6.6(b) relies on this, so the proof of graded non-isomorphism needs repair. The result is likely salvageable by a weight-set or corrected generator-count argument, but as written it is wrong.\n\nA second, less definite concern: the proof of Lemma 5.10 is load-bearing for Theorem 6.1 and is extremely hard to audit because the symbols for the three line types (through the origin, through (-1,-1), parallel through (-1,-1)) are not reliably distinguished in the preprint text. I suspect this is a typesetting issue rather than a mathematical gap, but the author should rewrite that proof with unambiguous notation. There are also small slips—for instance, a '=' in the proof of Proposition 5.8(b) that should be '≠'.\n\nBottom line: this paper deserves a serious referee. The central classification is new, meaningful, and mostly convincing; the isomorphism section is not in publishable shape. I would send it to review with a clear request to fix Lemma 6.5 and clarify Lemma 5.10. Once clean, I would cite Theorem 6.1 and would happily bring it to a reading group.","headline":"Main classification is solid and new, but the isomorphism section has a real generating-set error and the notation around Lemma 5.10 needs cleaning before this is publishable.","tokens_in":26561,"tokens_out":10888,"would_cite":true,"duration_ms":96464,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["17B30","17B66","17B65","17B05","17B70"],"pacs":[],"model":"deepseek-v4-flash","headline":"Every regular Borel subalgebra of Lie(Aut(A²)) has derived length 2 or 3: the only metabelian one is Lie(∆, ℓ(1,1)), and the length-3 ones are exactly Lie(t₂, ℓ(a,b), (a,b)) for a≠b.","keywords":["Borel subalgebra","regular subalgebra","derived length","solvable Lie algebra","homogeneous derivations","Lie algebra of vector fields","Aut(A²)","ind-group"],"falsifier":"Compute the derived series of Lie({∂ₚ,∂q,∂r}) for small triples of bidegrees, for example (a,b)=(2,0), (c,d)=(0,1), (e,f)=(1,0), using the determinant bracket formula. Lemma 5.9 predicts L⁽⁴⁾≠0 for certain configurations, and Lemma 5.10 predicts solvability only when one point dominates a line. If any small triple yields a non-Abelian solvable subalgebra not of the line-plus-point form, Theorem 6.1 fails. Also verify, for each listed length-3 B = Lie(t₂,ℓ(a,b),(a,b)) with a≠b, that B⁽²⁾≠0 and B⁽³⁾=0; any deviation would refute the theorem.","tokens_in":25450,"feed_emoji":"📐","tokens_out":16877,"duration_ms":149257,"temperature":0.7,"pith_summary":"This paper asks what the maximal solvable subalgebras of the Lie algebra Lie(Aut(A²)) look like when they are generated by homogeneous derivations with respect to the standard Z²-grading — the regular Borel subalgebras. The answer is a complete list: every such subalgebra has derived length 2 or 3, the only metabelian one is Lie(∆, ℓ(1,1)) (built from the diagonal ray of the bidegree lattice), and every derived-length-3 one is Lie(t₂, ℓ(a,b), (a,b)) for some (a,b) with a≠b, built from one homogeneous derivation plus all homogeneous derivations whose bidegrees lie on a single line through (−1,−1). The proof works by translating solvability into a combinatorial statement about sets of bidegrees: apart from the Abelian cases, a regular subalgebra is solvable exactly when all but one of its generators lie on such a line. Because maximal regular solvable subalgebras automatically are Borel, the list is complete; a corollary is that many of these Borel subalgebras do not arise from Borel subgroups of the automorphism group, so the classical correspondence between subgroups and subalgebras fails for this ind-group in infinitely many ways.","feed_headline":"A² automorphism Lie algebra: regular Borels have derived length 2 or 3","feed_subtitle":"One is metabelian; the rest are built from one lattice direction plus one extra derivation.","key_machinery":"Central object: the bracket [∂(a,b),∂(c,d)] = det[[c+1,a+1],[d+1,b+1]] ∂(a+c,b+d). It vanishes exactly when the two bidegrees lie on one line through (−1,−1). Lemma 5.10: Lie(S) is non-Abelian and solvable iff some (a,b)∈S, a≠b, has all other elements of S on ℓ(a,b). Lemma 3.7: every maximal regular solvable subalgebra contains t₂ and is automatically Borel; hence B = Lie(t₂,S), B^{(1)}=Lie(S), and the lemma forces S = ℓ(a,b)∩M ∪ {(a,b)}. Isomorphism classification uses the normal form G(n,1,λ,α,γ), where the integer n = gcd(a,b) is the only graded invariant.","core_discovery":"The central claim is Theorem 6.1: every regular Borel subalgebra of Lie(Aut(A²)) has derived length 2 or 3. The metabelian member is Lie(∆, ℓ(1,1)) = Lie(t₂,{(c,c): c≥1}), the span of the Euler field, ∂_{(0,0)}, and xyK[xy](x∂x − y∂y). All others are Lie(t₂, ℓ(a,b), (a,b)) for (a,b) ∈ Z²_{\\ge0} ∪ {(0,−1),(−1,0)} with a≠b: the torus t₂, the derivations along the ray ℓ(a,b)∩M, plus ∂_{(a,b)}. Each length-3 member is isomorphic as a Lie algebra to Lie(t₂, ℓ(0,−1), (a,0)) with a ∈ Z_{\\ge1}∪{−1}; the integer is gcd(a,b), and distinct members are not graded isomorphic. The member with a=−1 is infinitely generated, so only the positive-integer members could be pairwise abstractly isomorphic, which","pith_inferences":["Inference: The line-plus-point shape of S suggests that for Lie(Aut(Aⁿ)) the analogous regular Borel subalgebras would be controlled by flags of hyperplanes in the Zⁿ degree lattice rather than single lines; a concrete next computation would check whether the derived length of the n-dimensional analogue is n+1, which would beat the general 2n bound.","Inference: The unresolved question of abstract isomorphisms among Lie(t₂,ℓ(0,−1),(a,0)) for a≥1 can likely be settled by invariants such as the dimensions of the successive centralizers of the Euler element or the size of a minimal generating set; since the paper proves graded non-isomorphism only, non-graded invariants are the right tool.","Inference: The criterion can be tested computationally by enumerating small S⊂M and checking solvability via the derived series; this is a cheap verification of Lemma 5.10 and would harden the classification beyond the paper's case analysis."],"forward_implications":["Every homogeneous derivation is contained in an explicitly listed regular Borel subalgebra; the paper lists, for each bidegree, all regular Borel subalgebras containing that derivation.","The triangular subalgebras j⁺₂ and j⁻₂ are exactly the length-3 family members with (a,b) = (0,−1) and (−1,0), giving a new proof that they are Borel.","For every (c,d) ∈ Z²_{\\ge0} \\ {(0,0)}, the subalgebras Lie(∆,ℓ(1,1)) and Lie(t₂,ℓ(a,b),(a,b)) with (a,b) ∈ Z²_{\\ge0}, a≠b, do not correspond to any Borel subgroup of Aut(A²), so the subgroup–subalgebra correspondence fails in infinitely many ways.","Non-regular Borel subalgebras exist: any Borel subalgebra containing ∂_{(0,1)} + ∂_{(1,0)} is non-regular.","Up to isomorphism, the length-3 regular Borel subalgebras form the countable family Lie(t₂, ℓ(0,−1), (a,0)), a ∈ Z_{\\ge1}∪{−1}; the metabelian diagonal member is the sole length-2 representative and is not isomorphic to any other regular Borel subalgebra."],"fun_headline_variants":["Regular Borels of Aut(A²): derived length 2 or 3","All regular Borel subalgebras of Aut(A²) found: length 2 or 3","Regular Borels in Aut(A²): either metabelian or length 3","Derived length 2 or 3: regular Borel subalgebras of Aut(A²)","Aut(A²) regular Borels: only two derived lengths possible"],"cache_read_input_tokens":2304,"weakest_assumption_plain":"The classification collapses if Lemma 5.10 is false: the claim that every non-Abelian solvable regular subalgebra has all but one generator on a single line ℓ(a,b) through (−1,−1). The proof of that lemma is a long case analysis with easily confused line symbols (ℓ, ℓ, eℓ), so a missed configuration there would leave the list of regular Borel subalgebras incomplete.","fun_headline_variants_meta":{"raw":{"variants":["Regular Borels of Aut(A²): derived length 2 or 3","All regular Borel subalgebras of Aut(A²) found: length 2 or 3","Regular Borels in Aut(A²): either metabelian or length 3","Derived length 2 or 3: regular Borel subalgebras of Aut(A²)","Aut(A²) regular Borels: only two derived lengths possible"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000409,"raw_usage":{"total_tokens":1950,"prompt_tokens":731,"completion_tokens":1219,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":475,"completion_tokens_details":{"reasoning_tokens":1110}},"tokens_in":475,"tokens_out":1219,"duration_ms":8808,"temperature":1.0,"reasoning_tokens":1110,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-03T09:13:16.650787+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute the derived series of Lie({∂ₚ,∂q,∂r}) for small triples of bidegrees, for example (a,b)=(2,0), (c,d)=(0,1), (e,f)=(1,0), using the determinant bracket formula. Lemma 5.9 predicts L⁽⁴⁾≠0 for certain configurations, and Lemma 5.10 predicts solvability only when one point dominates a line. If any small triple yields a non-Abelian solvable subalgebra not of the line-plus-point form, Theorem 6.1 fails. Also verify, for each listed length-3 B = Lie(t₂,ℓ(a,b),(a,b)) with a≠b, that B⁽²⁾≠0 and B⁽³⁾=0; any deviation would refute the theorem.","supporting_citations":[],"review_version":1}