{"id":"f99cb7ca-c9e6-4108-8aa5-7b928685b5d1","arxiv_id":"2509.03742","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"For the handlebody Torelli groups HI and HBI, the kernels of the Johnson-detected cup product maps in second rational cohomology are computed explicitly as sums of irreducible SL_g(Q) modules.","lead":"A pure mathematics paper computes the second rational cohomology (the cup products) for two new Torelli-type subgroups of handlebody symmetry groups, using Johnson homomorphisms and representation theory. It gives the cup-product kernels as explicit lists of irreducible representations, complete for one group and up to one undecided summand for the other.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Exactness of Theorems A/B hinges on unverifiable highest-weight-vector and orbit computations in §12.4 and §7.4; one misidentified vector would shrink the claimed kernel.","rationale":"The paper's central achievement is an exact computation: Theorems A and B assert that the kernels of τ* and Θ* are precisely the specified SL_g(Q)-modules. The lower bounds are largely transparent (bracket map, J_2, τ-detected modules), but the upper bounds require eliminating every other irreducible constituent by exhibiting it in the image of the induced homology map. The method is standard—abelian cycles and highest weight vectors—but the elimination is only as strong as the explicit vector calculations. Several of these calculations are literally unreadable in the arXiv posting (§7.4, §12.4), and none are backed by machine-checked scripts. Because the argument invokes Schur's lemma, a single typographical error or misidentified weight could make a module erroneously survive, so the exactness of the kernel is not presently verifiable. The reader's verdict of CONDITIONAL is exactly right; flagging this does not move the verdict, but it identifies the precise repair needed. Independent support: the method is benchmarked against Hain's known kernel for Ig and Pettet's orbit claims, and conjectural parts are explicitly separated, so there is no reason to suspect the results are false; they are just not yet certifiable.","tokens_in":173384,"tokens_out":4276,"duration_ms":46753,"concrete_test":"Repair the corrupted formulas in §12.4 and run an independent LiE/Sage check for g=6: for each module in Λ²\\bar U_Q not listed in Theorem 11.1, verify the displayed vector is a highest weight vector (killed by E_ij for i<j) with the stated weight and lies in the SL_g-orbit of Lemma 12.2's Type 1/2/3 vectors. The most targeted single check is the vector v=Σ_{i≠j}x_ij+Σ_i(y_i+z_i) claimed for Φ_{0,...,0,1,0}: test E_ij v = 0 and compute its weight; if it fails, Theorem A's kernel is larger than stated. Similarly recompute -E_ik[(a_i∧a_j)⊗b_i]⊗[b_k⊗b_k] to verify Lemma 7.4 Type 3; if it does not equal the displayed Type 3 vector, the Theorem B upper bound needs revision.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The upper-bound half of Theorem A (Proposition 12.1) is the load-bearing step. To conclude that ker(τ*) is exactly Table 1, the proof must show that every SL_g(Q)-module in Λ²\\bar U*_Q not listed is dual to a module in im(τ*). The argument uses Lemma 12.2 to place Type 1–3 vectors in im(τ*), then §12.4 defines vectors x_ij, y_i, z_i, x_i, y claimed to be highest weight vectors for modules such as Φ_{0,...,0,1,0}, Φ_{0,...,0,1,0,1}, and Φ_{0,1,0,...,0,2}, and invokes Schur's lemma to conclude the whole module lies in im(τ*). In the posted text these definitions are corrupted into repeated unreadable glyph sequences; the 'N terms' expression in Lemma 12.2 is undefined as printed. The analogous upper-bound input for Theorem B, Lemma 7.4 (Types 3 and 4), is likewise garbled. No LiE script, character table, or independent certificate is supplied for the decompositions (Table 10 and §12.4), and Lemma 7.2's orbit statement is imported from Pettet without verification. If any displayed vector is not a highest weight vector of the stated weight, or is not in the claimed SL_g-orbit, then a module currently declared to be in im(τ*) could actually belong to ker(τ*), overcounting the kernel and invalidating the exact equality. This is a real verification gap, not an observed contradiction: the strategy is coherent and cross-checked against Hain's Γ_{0,2} description, but the decisive computations are not checkable from the manuscript.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper introduces two handlebody Torelli subgroups, HI^b_{g,p} (acting trivially on boundary homology) and HBI^b_{g,p} (acting trivially on handlebody homology), and studies the cup-product maps induced by the Johnson homomorphisms on their first rational cohomology. The main results are Theorem A, which gives an exact SL_g(Q)-module decomposition of ker(τ*: H^2(U;Q)→H^2(HI;Q)) for g≥6, and Theorem B, which gives an almost exact decomposition of ker(Θ*) for g≥3 up to one unresolved module. The proofs use five-term exact sequences, the Morita bracket, abelian cycles, and extensive highest-weight-vector computations in SL_g(Q) representations, with the HI_g result reformulated as ker(τ*) = Λ²Ubar*_Q ∩ (Γ_{0,2}+Q) in Remark 11.5.","tokens_in":173682,"tokens_out":3492,"duration_ms":46567,"significance":"If the computations are correct, this is a substantial contribution to the rational cohomology of handlebody Torelli groups, extending the Johnson–Hain–Morita–Pettet framework to these subgroups. The explicit lower-bound computations (e.g., evaluations of τ and J on bounding pair maps in Sections 6.2 and 6.3) are concrete and the comparison with Hain's theorem is a useful sanity check. However, the upper-bound arguments, which are load-bearing for the exact statements, are currently not verifiable from the posted text: several key vector definitions are corrupted and no computational certificates are supplied.","major_comments":[{"comment":"The proof of Proposition 12.1 depends on vectors x_ij, y_i, z_i, x_i, y claimed to be highest weight vectors for modules such as Φ_{0,...,0,1,0}, Φ_{0,...,0,1,0,1}, and Φ_{0,1,0,...,0,2}. In the posted text these definitions are corrupted into unreadable glyph sequences, and the 'N terms' expression in Lemma 12.2 is undefined. A single misidentified highest weight vector would change the kernel in Theorem A. The author must restore these definitions and provide an independent check (e.g., a LiE script or explicit character computation).","section":"§12.4 / Lemma 12.2"},{"comment":"Lemma 7.4 is the key input for the upper bound in Theorem B: Types 3 and 4 vectors are used in Proposition 7.3 to show that all but two modules in the ((Λ²V_Q)⊗V*_Q)⊗Sym²(V*_Q) summand lie in im(Θ*). The proof of Lemma 7.4 is garbled after the first displayed calculation, so the claimed sl_g(Q) actions producing Type 3 and Type 4 vectors cannot be checked. This affects the exactness of the two-alternative description in Theorem B.","section":"§7.4 / Lemma 7.4"},{"comment":"The module decompositions in Tables 3–10 and the branching computations in §12.4 are load-bearing but are asserted without reproducible scripts. The text refers to LiE in Appendix D, but no scripts, log files, or character certificates are included. Without these, a referee cannot independently confirm the multiplicities or the claimed highest weight vectors. Please provide the computational certificates or a fully explicit verification.","section":"Tables 3–10 and §12.4"},{"comment":"The membership of Φ_{0,1,0,...,0,1} (g≥4) or Φ_{0,2} (g=3) in ker(Θ*) is left undecided, and the theorem is stated as a disjunction. This is honest, but it means the description of ker(Θ*) is not complete. If this is the intended contribution, the unresolved case should be highlighted as an explicit open problem in the introduction and again in the conclusion.","section":"Theorem B / Conjecture 1.2"}],"minor_comments":[{"comment":"The abstract says the paper 'describes cup products' in H^2, but the theorems only determine the kernels of τ* and Θ*. The full cup product description is conditional on Conjectures 1.1 and 1.2. The wording should be sharpened to avoid overstatement.","section":"Abstract and §1.4"},{"comment":"The claim that Table 10 'stabilizes when g≥6' is not justified beyond the table. A sentence indicating the Weyl character formula or a character computation would help.","section":"§10.1"},{"comment":"The orbit statements of Lemmas 6.5 and 7.2 are quoted from Pettet. Since they are essential for the upper bounds, the text should give precise references to the corresponding statements, and ideally reproduce the arguments in an appendix if they are not easily accessible.","section":"§6.5 and §7.2"},{"comment":"The notation p+b=1 and the separate cases p=1,b=0 and p=0,b=1 are used interchangeably in places (e.g., 'p=1 and b=1' in §1.7). Please standardize this.","section":"Notation"}],"recommendation":"major_revision","confidential_remarks":"The paper is plausible and the lower-bound arguments are mostly explicit, but the current text is not checkable in the places where the exactness of the main theorems is decided. I would not reject outright, but the author must make §7.4 and §12.4 readable and provide computational certificates or detailed highest-weight-vector verifications before the paper can be accepted."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"I read Holden's paper on Johnson homomorphisms and the second rational cohomology of handlebody Torelli groups. Short version: this is a real piece of work, but as posted it is not checkable in its load-bearing sections, and the gap is exactly where the upper-bound arguments live.\n\nWhat's new: the groups HI^b and HBI^b are clearly defined, and Theorems A and B give precise SL_g(Q)-decompositions of ker(tau*) and ker(Theta*) — exact for HI_g when g>=6, two-possibility for HBI_g when g>=3. That is a genuine extension of the Hain–Morita–Pettet–Sakasai program to handlebody Torelli groups. The paper is honest: it separates proved results from Conjectures 1.1–1.3, and it benchmarks against Hain's Γ_{0,2}+Q description rather than adjusting definitions to fit. The lower-bound arguments are written out in detail, and the evaluation of tau and J on explicit bounding pair maps in Sections 6.2 and 6.3 is the kind of concrete computation that makes the paper useful.\n\nSoft spots, in proportion. The stress-test note is on target: the upper-bound halves of Theorems A and B hinge on Section 12.4's highest-weight-vector constructions and Lemma 7.4's Type 3/4 vectors, and in the posted text those passages are corrupted into unreadable glyph sequences. Lemma 12.2's \"N terms\" is undefined as printed. No LiE scripts or character certificates back Tables 3–10 or the module orbit claims. If any of those displayed vectors is not a highest weight vector of the stated weight, or not in the claimed orbit, a module currently declared to lie in im(tau*) would actually belong to ker(tau*), overcounting the kernel and invalidating the exact equality. That is a real verification gap. I want to be clear: it is not an observed contradiction. The strategy is coherent and the external check against Hain's theorem is reassuring; but a referee cannot, as things stand, certify the upper bound.\n\nAlso minor but real: Theorem B is genuinely two-valued. The deciding containment of Φ_{0,1,0,...,0,1} (g>=4) or Φ_{0,2} (g=3) in ker(Theta*) is flagged as conjectural, so the headline cup-product statement for HBI is incomplete by the authors' own admission. Sections 13–15 were outside my review budget, so the decorated cases rest on the stated spectral sequence structure rather than checked details.\n\nWho is this for: anyone working on Torelli groups, Johnson homomorphisms, or low-dimensional cohomology of mapping-class-like groups. It deserves a serious referee, but the referee report should be conditional: repair the corrupted formulas, supply the LiE inputs and certificates, and if possible resolve the undecided module. I would not desk-reject it.\n\nRecommendation: send to peer review, with the expectation of revision. Cite it if you work in this area, but only after the computations are checkable.","headline":"A solid, honestly delimited computation in the Hain–Morita tradition that deserves refereeing, but the posted text garbles exactly the steps a referee needs to check.","tokens_in":174370,"tokens_out":2728,"would_cite":true,"duration_ms":32463,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["57K20","20J06","17B10"],"pacs":[],"model":"deepseek-v4-flash","headline":"Kernel of the Johnson cup-product map is exactly eight modules for genus at least 6.","keywords":["handlebody group","Torelli group","Johnson homomorphism","cup product","group cohomology","SL_g representations","abelian cycles","mapping class group"],"falsifier":"Recompute Tables 3-10 and the vector identities in Sections 7.4 and 12.4 with an independent computer algebra system for g=6, and for g=3,4,5 where Theorem B applies, checking each claimed highest-weight vector against the action of tau* or Theta*. If any module asserted to lie in the image of tau* or Theta* actually has zero image, or if an imported orbit statement omits a module, the exact kernel descriptions fail. For Theorem B, evaluate the second Johnson homomorphism on the commutator of the two normal generators of HBI; the resulting module decides which of the two alternatives holds.","tokens_in":173076,"feed_emoji":"","tokens_out":8474,"duration_ms":97545,"temperature":0.7,"pith_summary":"Two Torelli subgroups of the handlebody group are introduced: HI, whose elements act trivially on the boundary surface's first homology, and HBI, whose elements act trivially on the handlebody's first homology. The paper restricts the classical Johnson homomorphisms to these subgroups and studies the induced cup-product maps from the second exterior power of the Johnson-image abelian quotient into second rational cohomology. Theorem A shows that for genus at least 6 the kernel of tau* for HI_g, HI_{g,1}, and HI^1_g is exactly the direct sum of eight explicit irreducible SL_g(Q)-modules, with only small decoration-dependent multiplicities. Theorem B shows that for genus at least 3 the kernel of Theta* for HBI is the short list in Table 2, with at most one further module whose membership remains conjectural. If the paper's two conjectural H^1 identifications hold, these tables describe the full rational cup product structure, not just the Johnson-detected part.","feed_headline":"Johnson cup-product kernel is exactly eight modules for genus at least 6","feed_subtitle":"Handlebody Torelli groups: the degree-2 classes seen by Johnson homomorphisms are now listed, up to two conjectures.","key_machinery":"The load-bearing object is the cup-product map viewed as an SL_g(Q)-equivariant map out of the second exterior power of the dual Johnson image: tau*: H^2(U;Q) -> H^2(HI;Q) and Theta*: H^2(W;Q) -> H^2(HBI;Q). Because SL_g(Q) representations decompose into irreducibles Phi_{w_1,...,w_{g-1}}, the kernel is a subrepresentation, so the problem reduces to deciding module by module whether the dual module lies in the cokernel. The cokernel is computed from the five-term exact sequence of 1 -> ker(Johnson) -> G -> image -> 1, whose boundary map b sends x wedge y to a commutator class; then the second Johnson homomorphism and the bracket map on Johnson images detect modules in the cokernel, while exp","core_discovery":"The paper's central claim is that the Johnson-detected part of the second rational cohomology of handlebody Torelli groups is governed by the representation theory of SL_g(Q). Restricting the Johnson homomorphisms of the mapping class group and of Aut(F_g) to the two handlebody Torelli groups gives abelian quotients U and W, and the induced cohomology maps factor through the second exterior powers of their duals. Theorem A (g >= 6) states that ker(tau*) is exactly Q + Phi_{0,1,0,...,0,1,0} + Phi_{0,2,0,...,0} + Phi_{1,0,...,0,1} + Phi_{1,0,...,0,1,1} + Phi_{1,1,0,...,0,1} + Phi_{2,0,...,0} + Phi_{2,0,...,0,2}, equivalently Lambda^2 Ubar*_Q intersect (Gamma_{0,2} + Q). Theorem B (g >= 3) stat","pith_inferences":["I would audit Sections 7.4 and 12.4 first: the theorems are exactly as strong as the vector identities there, and the posted text corrupts several of those computations into unreadable glyph sequences, so an independent recomputation is the quickest way to test the upper bounds.","The unresolved extra module in Theorem B should be checkable directly: evaluate the second Johnson homomorphism on the commutator of the two normal generators supplied for HBI; whichever module appears decides between the two alternatives.","The mixed summand ((Lambda^2 V_Q) tensor V*_Q) tensor Sym^2(V*_Q) is the genuinely new source of modules in Theorem B, since other Torelli-type groups do not have a symmetric-square piece from the symplectic image; these tables indicate what changes when an abelian radical is present.","If the stabilised decompositions hold for all larger genera, the same five-term-sequence plus abelian-cycle method should produce finite, genus-independent presentations for higher Johnson-image cohomology of these groups."],"forward_implications":["For genus at least 6, the eight-module list in Theorem A is complete for HI_g, HI_{g,1}, and HI^1_g, so every degree-2 class whose Johnson wedge is not one of those modules survives nontrivially in cohomology.","For HBI in genus at least 3, the kernel is pinned down except for one possible module; settling the conjectured membership of Phi_{0,2} or Phi_{0,1,0,...,0,1} would make Theorem B unconditional.","Under Conjectures 1.1 and 1.2, the entire cup product on first rational cohomology of these groups is determined by the tables, not only the Johnson-detected part.","The new SL_g(Q)-module decomposition of the second Johnson image of the handlebody Johnson kernel gives a conceptual reformulation of that image for all p+b at most 1.","The decoration cases differ only by known multiplicities of a few low-weight modules, so the high-genus structure is stable across closed, once-punctured, and once-bordered surfaces."],"supporting_citations":[{"why":"Introduces the Johnson homomorphism tau on the Torelli groups; its restriction defines the abelian quotients U and Ubar studied in Theorem A.","marker":"[11]"},{"why":"Establishes that tau captures the rational abelianization of the Torelli group, motivating the use of tau as the detected subspace of H^1.","marker":"[13, 14, 15]"},{"why":"Provides the baseline kernel computation for the full Torelli group that the paper adapts to the handlebody subgroups.","marker":"[18]"},{"why":"Supplies the imported orbit statements, Lemmas 6.5 and 7.2, used to show most modules in the decomposition lie outside the kernels.","marker":"[30]"},{"why":"Gives the normal generating sets for the two handlebody Torelli groups; the explicit annulus twists and disk twists used as abelian cycles come from these generators.","marker":"[29]"},{"why":"Computes the symplectic image of the handlebody group, yielding the block matrix form and the SL_g(Q)-action on U and W.","marker":"[20]"},{"why":"Introduces the bracket map on higher Johnson images, the main tool for detecting the lower-bound modules in ker(tau*).","marker":"[24]"},{"why":"Supplies the abelian-cycle method for realizing modules in the image of the induced homology map, used throughout the upper-bound arguments.","marker":"[33]"},{"why":"Determines the second Johnson image of the handlebody Johnson kernel as a trace-map kernel; Theorem 15.1 reformulates it as an SL_g(Q)-module decomposition.","marker":"[5]"}],"fun_headline_variants":["Johnson cohomology kernel for handlebody Torelli: exactly 8 modules (g≥6)","Cup-product kernel in handlebody Torelli: 8 modules, genus≥6","Handlebody Torelli: Johnson kernel is 8 explicit modules for g≥6","Eight modules determine Johnson cup product for handlebody Torelli (g≥6)","Johnson cohomology of handlebody Torelli: 8 modules, up to 2 conjectures (g≥6)"],"cache_read_input_tokens":2688,"weakest_assumption_plain":"The computation that each listed highest-weight vector really lies in the claimed SL_g(Q)-orbit, together with the imported orbit statements for the second Johnson map, is the load-bearing step; a single wrong vector identity could place a module inside a kernel where the theorem puts it outside.","fun_headline_variants_meta":{"raw":{"variants":["Johnson cohomology kernel for handlebody Torelli: exactly 8 modules (g≥6)","Cup-product kernel in handlebody Torelli: 8 modules, genus≥6","Handlebody Torelli: Johnson kernel is 8 explicit modules for g≥6","Eight modules determine Johnson cup product for handlebody Torelli (g≥6)","Johnson cohomology of handlebody Torelli: 8 modules, up to 2 conjectures (g≥6)"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001345,"raw_usage":{"total_tokens":5318,"prompt_tokens":776,"completion_tokens":4542,"prompt_tokens_details":{"cached_tokens":256},"prompt_cache_hit_tokens":256,"prompt_cache_miss_tokens":520,"completion_tokens_details":{"reasoning_tokens":4420}},"tokens_in":520,"tokens_out":4542,"duration_ms":33030,"temperature":1.0,"reasoning_tokens":4420,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-05T10:43:58.829160+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Recompute Tables 3-10 and the vector identities in Sections 7.4 and 12.4 with an independent computer algebra system for g=6, and for g=3,4,5 where Theorem B applies, checking each claimed highest-weight vector against the action of tau* or Theta*. If any module asserted to lie in the image of tau* or Theta* actually has zero image, or if an imported orbit statement omits a module, the exact kernel descriptions fail. For Theorem B, evaluate the second Johnson homomorphism on the commutator of the two normal generators of HBI; the resulting module decides which of the two alternatives holds.","supporting_citations":[{"cited_title":"Johnson, An abelian quotient of the mapping class group Ig, Math","cited_arxiv_id":null,"evidence_quote":"Introduces the Johnson homomorphism tau on the Torelli groups; its restriction defines the abelian quotients U and Ubar studied in Theorem A."},{"cited_title":"The Johnson homomorphism and the second cohomology of IA_n","cited_arxiv_id":"math/0501053","evidence_quote":"Supplies the imported orbit statements, Lemmas 6.5 and 7.2, used to show most modules in the decomposition lie outside the kernels."},{"cited_title":"Hirose, The action of the handlebody group on the first homology group of the surface, Kyungpook Math","cited_arxiv_id":null,"evidence_quote":"Computes the symplectic image of the handlebody group, yielding the block matrix form and the SL_g(Q)-action on U and W."},{"cited_title":"Morita, Abelian quotients of subgroups of the mapping class group of surfaces, Duke Math","cited_arxiv_id":null,"evidence_quote":"Introduces the bracket map on higher Johnson images, the main tool for detecting the lower-bound modules in ker(tau*)."},{"cited_title":"Sakasai, The Johnson homomorphism and the third rational cohomology group of the Torelli group, Topology Appl., 148 (2005), 83-111","cited_arxiv_id":null,"evidence_quote":"Supplies the abelian-cycle method for realizing modules in the image of the induced homology map, used throughout the upper-bound arguments."},{"cited_title":"The handlebody group and the images of the second Johnson homomorphism","cited_arxiv_id":"2010.16268","evidence_quote":"Determines the second Johnson image of the handlebody Johnson kernel as a trace-map kernel; Theorem 15.1 reformulates it as an SL_g(Q)-module decomposition."}],"review_version":1}