{"id":"c1b37c6b-c25d-4497-bb39-c2234275865a","arxiv_id":"1908.04608","paper_version":2,"verdict":"ACCEPT","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"low","formal_verification":"none","parameter_count":0,"one_line_summary":"Parafree Lie algebras over Z/2 with nonzero second homology are constructed explicitly; the Z-completion of the free rank-two Lie algebra has cohomological dimension greater than two; countable parafree groups with H2 different from zero exist.","lead":"This paper constructs explicit infinite Lie algebras that look free at every finite level but still have nonzero second homology, and shows the completed free Lie algebra over Z has cohomological dimension above two. It also proves, non-constructively, that countable parafree groups with nonzero second homology exist.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Theorem B's homology class is not defined as written: §2.3's s-definition fails the cycle condition, and the infinite sum requires an unproved completed-exterior-square isomorphism.","rationale":"The reader correctly identified the missing topology/completed-exterior-square justification as a weakness. My stress-test finds a more basic problem in the same passages: the displayed definition of s_{2^n} in §2.3 is inconsistent with the identity quoted in §2.1, and for n = 2 the resulting t_4 fails to be a cycle. This is not a matter of filling in a routine detail; as written, the central homology class of Theorem B is not defined. The claim may be repairable by correcting the upper limit and adding the missing pro-object argument, which is why a flat REJECT is not warranted, but the paper should not be accepted without those fixes. Hence the verdict should move from ACCEPT to CONDITIONAL.","tokens_in":12016,"tokens_out":33514,"duration_ms":282408,"concrete_test":"Use a Hall basis, or the embedding of f_Z(a,b) into the free associative algebra, to compute [r_4, a] + [s_4, b] with the §2.3 definition: r_4 = [a, 4b], s_4 = −[[a, 3b], a] + [[a, 2b], [a, b]] − [[a, b], [a, 2b]]. If the result is nonzero, t_4 is not a cycle and the §2.3 sum does not define an H2 element. If the published upper limit is actually n (matching §2.1), repeat the computation with that s; if the cycle condition then holds, the remaining check is whether a completed exterior square and the isomorphism H2(hat g) ≅ Ker(hat g ∧̂ hat g → hat g) can be proved for these infinite sums.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The construction of the nonzero homology class in Theorem B (and the corresponding part of Theorem 2.11) has a concrete algebraic gap before the topological one. In §2.3 the element s_{2^n} is displayed with upper limit 2^n−1, while the identity quoted from [13, Lemma 4.1] in §2.1, and the definition used there, are compatible only with upper limit n. For the §2.3 definition, t_{2^n} = r_{2^n} ∧ a + s_{2^n} ∧ b is not a cycle. For n = 2 (that is, 2^n = 4), a direct Jacobi expansion gives [r_4, a] + [s_4, b] = [[a, 3b], [a, b]] ≠ 0 in the free Lie algebra f_Z(a,b), so t_4 is not in Ker(f ∧ f → f) and the infinite sum ∑ 2^n t_{2^n} cannot define an element of H2(hat f, Z) by Proposition 2.2. Even if the upper limit is corrected to n, the proof still treats infinite sums such as ∑ 2^n t_{2^n} as elements of an exterior square of the pronilpotent completion. Proposition 2.2 is stated for discrete Lie algebras, and the paper neither defines a completed exterior square nor proves the needed isomorphism H2(hat g) ≅ Ker(hat g ∧̂ hat g → hat g). Thus the nonzero-homology conclusions in Theorem 2.11 and Theorem B rest on an invalid cycle condition and an unproved interchange of homology with inverse limits.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies homological properties of parafree Lie algebras. In Theorem A it constructs an explicit countable parafree Lie algebra over a field of characteristic 2 with nonzero second homology. In Theorem B it claims that the pronilpotent completion of the free Lie algebra of rank two over Z has cohomological dimension at least 3, by constructing a nonzero 2-divisible element in H2. Section 2.4 proves existence of countable parafree groups with nonzero H2, following Bousfield's theorem. The proofs use a Lie-algebra analog of Stallings' theorem, a Lie analog of the lamplighter group, and non-rationality of lacunary power series.","tokens_in":12280,"tokens_out":9764,"duration_ms":83174,"significance":"If the proofs are completed, Theorem A would give the first explicit example of a parafree Lie algebra with nontrivial second homology, and Theorem B would settle a Lie-algebra counterpart of a question about cohomological dimension of pronilpotent completions. The group-theoretic result in Section 2.4 is also new. The overall approach is original, and the idea of detecting nonzero homology via projections to a lamplighter Lie algebra and non-rational power series is elegant. The paper is clearly structured and mostly self-contained, with useful references to prior work.","major_comments":[{"comment":"The displayed definition of s_{2^n} in the proof of Theorem B has upper summation limit 2^n-1, whereas the identity quoted from [13, Lemma 4.1] and the definition used in Theorem 2.11 require upper limit n. With the printed upper limit the element t_{2^n} is not a cycle: for n=2, a direct Jacobi expansion gives [r_4,a]+[s_4,b] = [[a,3b],[a,b]] ≠ 0 in the free Lie algebra f_Z(a,b). Therefore the infinite sum ∑ 2^n t_{2^n} cannot be claimed to represent an element of H2(ĥf,Z) via Proposition 2.2. This is a load-bearing error for Theorem B and must be corrected, presumably by setting the upper limit to n, and the cycle condition re-verified.","section":"Section 2.3, definition of s_{2^n}"},{"comment":"The proofs of Theorem 2.11 and Theorem B treat infinite sums such as R_α = ∑ α_n r_{2^n}, S_α = ∑ α_n s_{2^n}, and ∑ 2^n t_{2^n} as elements of the exterior square ĥf ∧ ĥf and then as cycles in Ker(ĥf ∧ ĥf → ĥf). However, Proposition 2.2 is stated for the algebraic exterior square of a discrete Lie algebra, and the algebraic exterior square does not contain arbitrary infinite sums. The paper neither defines a completed exterior square nor proves the natural isomorphism H2(ĥg) ≅ Ker(ĥg ∧̂ ĥg → ĥg) for pronilpotent completions. Without such a result, the nonzero-homology conclusions in Theorem 2.11 and Theorem B do not follow from the written arguments. The authors should either define the completed construction and establish the isomorphism, or provide an alternative argument via finite truncations and inverse limits.","section":"Section 2.1, Theorem 2.11 and Section 2.3, Theorem B"}],"minor_comments":[{"comment":"The expression '2^{n-i}b' is printed as '2n−ib' in several places, which is ambiguous; it should be typeset as a power of 2.","section":"Notation throughout"},{"comment":"The equality ∑ 2^n t_{2^n} = 2^k ∑_{n≥k+1} 2^{n-k} t_{2^n} in H2(ĥf,Z) relies on the same unproved completed-exterior-square identification; after the main gap is fixed, the authors should also justify that the infinite tail again defines a homology class.","section":"Section 2.3, 2-divisibility argument"},{"comment":"The argument that r ∩ [e,e] = [r,e] in the proof of Lemma 2.13 is compact; a sentence explaining why the K-linear span of the α_i, β_i contributes nothing to r ∩ [e,e] would improve readability.","section":"Section 2.2, proof of Lemma 2.13"}],"recommendation":"major_revision","confidential_remarks":"The main novelty of the paper appears to be Theorem A, which is not affected by the completed-exterior-square gap. Theorem B is more ambitious and needs substantive repair. The issues identified are technical in nature and seem fixable within the manuscript's scope, so I do not recommend rejection. The paper cites its own earlier work appropriately for the quoted identity, and the group-theoretic result in Section 2.4 is sound."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague —\n\nQuick take: the paper has two genuinely useful results and one load-bearing gap. Theorem A (an explicit countable parafree Lie algebra over a field of characteristic two with H2 ≠ 0) and Proposition 1 (countable parafree groups with nonzero H2) are new and, as far as I can tell, correct. Theorem B, the claim that the pronilpotent completion of the free Lie algebra over Z has cohomological dimension > 2, is not proved as written.\n\nThe construction of the lamplighter Lie algebra l_R and the projection from the free Lie algebra to it is the right adaptation of the group method. The non-rationality lemmas (2.8–2.10) are clean, and the Lie-algebra Stallings theorem is applied sensibly. Theorem A avoids the problem I’m about to describe because the infinite series live in the Lie algebra itself; only brackets are computed there, and convergence is harmless. The filtered-union argument for groups is short and correct.\n\nThe soft spot is §2.3. Two things go wrong. First, the displayed definition of s_{2^n} has upper limit 2^n−1, which is a typo; the cited identity from [13, Lemma 4.1] requires upper limit n, and with the printed upper limit t_4 is not a cycle (the stress-test’s Jacobi computation checks out). That alone is minor. Second, and not minor: the proof writes the infinite sum ∑ 2^n t_{2^n} as a cycle in the exterior square of the completion. The paper never puts a topology on f∧f, never defines a completed exterior square, and never proves H_2(hat f) ≅ Ker(hat f ∧̂ hat f → hat f). The ordinary exterior square of the completion has no infinite linear combinations, so the displayed sum simply is not an element of it. Theorem B’s nonzero class is exactly this sum, so the cohomological-dimension conclusion does not follow. The 2-divisibility paragraph shares the same problem.\n\nThis is probably fixable: one can likely prove a Mittag-Leffler type statement or use the Bousfield spectral sequence to identify the homology. But the paper as it stands is not complete.\n\nWho gets value: researchers on parafree groups and Lie algebras, and anyone working on pro-nilpotent completions. It deserves a proper peer review, but a serious referee should send it back for a rewritten §2.3.","headline":"Two of the three main results hold up; Theorem B's proof has a real gap in how it treats the completed exterior square.","tokens_in":12894,"tokens_out":11823,"would_cite":true,"duration_ms":114078,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["17B55","17B01","17B65"],"pacs":[],"model":"deepseek-v4-flash","headline":"Explicit parafree Lie algebra has nonzero second homology","keywords":["parafree Lie algebra","second homology","pronilpotent completion","cohomological dimension","formal power series","non-rational series","lamplighter Lie algebra","Stallings theorem"],"falsifier":"Compute inside the completed exterior square \\hat f \\tilde\\wedge \\hat f and check whether the partial sums \\sum_{n=1}^N 2^n (r_{2^n} ∧ a + s_{2^n} ∧ b) form a Cauchy sequence for any pronilpotent topology; if in the limit the sum is zero because the filtration of the exterior square is not Hausdorff, then the claimed 2-divisible class in H2(\\hat f, Z) vanishes and Theorem B collapses.","tokens_in":11736,"feed_emoji":"🧮","tokens_out":10921,"duration_ms":95187,"temperature":0.7,"pith_summary":"The paper constructs an explicit countable parafree Lie algebra with nonzero second homology, the first such example given by generators and relations. Over a field K of characteristic two, the quotient a = b/γω(b) of the explicitly presented algebra b is parafree, and it carries a nonzero class in H2(a,K). The same method shows that the pronilpotent completion of the free Lie algebra on two generators over Z has cohomological dimension greater than two, because H2 of that completion contains a nonzero element divisible by every power of two. The paper also proves that the pronilpotent completion of a free group of rank at least two contains uncountably many countable parafree subgroups with nonzero H2. A sympathetic reader should care because it pins down where the Parafree Conjecture can hold: finite generation is essential, and explicit homological counterexamples exist just beyond it.","feed_headline":"Explicit parafree Lie algebra has nonzero second homology","feed_subtitle":"In countable generation, parafree no longer forces trivial homology; free completions can have dimension > 2.","key_machinery":"The central object is the Lie-algebra analogue of the lamplighter group: l_R = R[x] ⋊ Q, where Q is a one-dimensional free module R t and the generator t acts on polynomials by [p,t] = px; its pronilpotent completion is R[[x]] ⋊ Q with [f,t] = fx. Working through the Lyndon–Hochschild–Serre spectral sequence, the paper identifies H2(\\hat l_R, R) with the tensor product R[[x]] \\wedge_{R[x]} R[[x]] over the polynomial ring (Lemmas 2.6–2.7), and shows that a class p ∧ 1 is nonzero whenever p is a formal power series that is not a rational function (Lemmas 2.8–2.9). The cycles feeding this detection are produced in the free Lie algebra f_R(a,b) by the Engel-type identity [ [x, 2^n y], x] = [ \\sum_{i=0}^{n-1} (-1)^i [[x, $2^{{n-1-i}}$ y], [x, 2^i y]], y ], which makes r_{2^n} = [a, 2^n b] and s_{2^n} = \\sum_{i=1}^n (-1)^i [[a, $2^{{n-i}}$ b], [a, $2^{{i-1}}$ b]] satisfy [r_{2^n}, a] + [s_{2^n}, b] = 0; hence weighted infinite sums R_α ∧ a + S_α ∧ b are cycles in \\hat f \\tilde\\wedge \\hat f. Their images in H2(\\hat l_R) are non-rational series p_α ∧ 1 = (\\sum α_n $x^{{2^n}}$) ∧ 1, establishing uncountability in Theorem 2.11, and the power series \\sum 2^n $x^{{2^n}}$ ∧ 1 produces the 2-divisible class for Theorem B. For Theorem A, a characteristic-2 identity, [a, $2^{{n+1}}$ b, a] = [a, 2^n b, a, 2^n b], converts the defining relations of b into the same non-rationality argument and yields x_1 ∧ a + y_1 ∧ b as the nonzero class.","core_discovery":"Over a field K of characteristic 2, let b be the Lie algebra generated by a, b, and two countable families {x_i}_{i≥1}, {y_i}_{i≥1} subject to x_i = [a,b,b] + [x_{i+1}, 2^i b] and y_i = [a,b,a] + [y_{i+1}, 2^i b] for all i ≥ 1. The paper proves that a = b/γω(b), where γω(b) is the intersection of the lower central series, is parafree: the map from the free Lie algebra on {a,b} sending the generators to a and b induces isomorphisms on every lower central quotient. It then shows H2(a,K) ≠ 0 by exhibiting the cycle x_1 ∧ a + y_1 ∧ b and proving its image in the homology of a related lamplighter Lie algebra is nonzero. For the integral case, writing f for the free Lie algebra on two generators over Z and \\hat f for its pronilpotent completion, the paper proves that \\sum_{n≥1} 2^n (r_{2^n} ∧ a + s_{2^n} ∧ b) defines a nonzero element of H2(\\hat f, Z), that this element is 2^k-divisible for every k, and consequently that the cohomological dimension of \\hat f is at least 3. Finally, it shows that the pronilpotent completion of a noncyclic finitely generated free group is the filtered union of its countable parafree subgroups, and since H2 commutes with filtered colimits, uncountably many of those countable parafree groups have nonzero second homology.","pith_inferences":["The same non-rationality technique should produce finitely presented parafree Lie algebras with nonzero H2 over Z, not just countably presented ones: characteristic 2 is convenient for the Engel identities, but the 2-divisibility argument of Theorem B is characteristic-free and could be sharpened.","The method suggests a route to explicit countable parafree groups with nonzero H2: if a group homomorphism from a free group to a lamplighter-type group can be arranged so that a non-rational series arises from a cycle in the group's exterior square, the existence proof via HZ-localization could be replaced by concrete relations.","A natural testable extension is whether H2(\\hat f, Z) contains a copy of Z[1/2] rather than just one 2-divisible element; the paper does not address the full structure of this homology group."],"forward_implications":["The Parafree Conjecture as usually stated for finitely generated groups has no direct analogue for countably generated Lie algebras: Theorem A supplies an explicit countable parafree Lie algebra with H2 ≠ 0.","The pronilpotent completion of a free Lie algebra of rank at least two over Z has cohomological dimension at least three, so Stallings–Swan-type dimension-one rigidity fails in the category of pronilpotent Lie algebras.","H2(\\hat f, Z) contains a nonzero element divisible by every power of 2, so it is not a free abelian group; consequently the trivial module over U(\\hat f) has projective dimension at least three.","There are uncountably many countable parafree subgroups of the pronilpotent completion of a free group with nonzero H2, showing the countable parafree landscape is large rather than a single pathology.","The explicit cycles are detected by formal power series that are not rational, so any parafree Lie algebra admitting a similar map to the lamplighter Lie algebra with non-rational image will again have nonzero H2."],"supporting_citations":[{"why":"Gives the exterior-square description H2(g,R) ≅ Ker(g ˜∧ g → g) used to name explicit cycles.","marker":"[12]"},{"why":"Supplies the method of detecting nonzero homology via non-rational power series, and the Engel identity used to build cycles in the free pronilpotent completion.","marker":"[13]"},{"why":"Provides the group-theoretic construction of divisible homology classes that Theorem B adapts to Lie algebra completions.","marker":"[14]"},{"why":"Stallings' theorem, in its Lie algebra version, is used to prove parafreeness of a from a 2-connected map.","marker":"[17]"},{"why":"Bousfield's result that H2 of the pronilpotent completion of a free group is uncountable motivates Theorems 2.11 and the group-theoretic Proposition 1.","marker":"[6]"},{"why":"Gives the Γ-system-of-equations and HZ-localization machinery that proves every countable subset of the free group completion lies in a countable parafree subgroup.","marker":"[15]"},{"why":"Defines parafree Lie algebras and frames the question of which homological properties they share with free Lie algebras.","marker":"[5]"}],"fun_headline_variants":["Countable parafree Lie algebra breaks trivial-homology rule","Explicit parafree Lie algebra with nonzero H_2 over F_2","Pronilpotent completion dimension leaps beyond 2","Countable parafree group with nontrivial H_2","Parafree no longer implies trivial second homology"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that infinite linear combinations such as \\sum α_n (r_{2^n} ∧ a + s_{2^n} ∧ b) and \\sum 2^n (r_{2^n} ∧ a + s_{2^n} ∧ b) are legitimate elements of the exterior square of the pronilpotent completion, and that H2 of the completion still equals the kernel of the wedge map; the paper does not define the topology or completed exterior square making these sums converge.","fun_headline_variants_meta":{"raw":{"variants":["Countable parafree Lie algebra breaks trivial-homology rule","Explicit parafree Lie algebra with nonzero H_2 over F_2","Pronilpotent completion dimension leaps beyond 2","Countable parafree group with nontrivial H_2","Parafree no longer implies trivial second homology"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001329,"raw_usage":{"total_tokens":5412,"prompt_tokens":956,"completion_tokens":4456,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":572,"completion_tokens_details":{"reasoning_tokens":4369}},"tokens_in":572,"tokens_out":4456,"duration_ms":30246,"temperature":1.0,"reasoning_tokens":4369,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T13:39:31.918392+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute inside the completed exterior square \\hat f \\tilde\\wedge \\hat f and check whether the partial sums \\sum_{n=1}^N 2^n (r_{2^n} ∧ a + s_{2^n} ∧ b) form a Cauchy sequence for any pronilpotent topology; if in the limit the sum is zero because the filtration of the exterior square is not Hausdorff, then the claimed 2-divisible class in H2(\\hat f, Z) vanishes and Theorem B collapses.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Gives the exterior-square description H2(g,R) ≅ Ker(g ˜∧ g → g) used to name explicit cycles."},{"cited_title":"Ivanov, R","cited_arxiv_id":null,"evidence_quote":"Supplies the method of detecting nonzero homology via non-rational power series, and the Engel identity used to build cycles in the free pronilpotent completion."},{"cited_title":"Ivanov, R","cited_arxiv_id":null,"evidence_quote":"Provides the group-theoretic construction of divisible homology classes that Theorem B adapts to Lie algebra completions."},{"cited_title":"Stallings: Homology and central series of groups, J","cited_arxiv_id":null,"evidence_quote":"Stallings' theorem, in its Lie algebra version, is used to prove parafreeness of a from a 2-connected map."},{"cited_title":"Bousfield: Homological localization towers for gro ups and π -modules, Mem","cited_arxiv_id":null,"evidence_quote":"Bousfield's result that H2 of the pronilpotent completion of a free group is uncountable motivates Theorems 2.11 and the group-theoretic Proposition 1."},{"cited_title":"Farjoun, K","cited_arxiv_id":null,"evidence_quote":"Gives the Γ-system-of-equations and HZ-localization machinery that proves every countable subset of the free group completion lies in a countable parafree subgroup."},{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"Defines parafree Lie algebras and frames the question of which homological properties they share with free Lie algebras."}],"review_version":1}