{"id":"262a7b9e-d856-4def-961d-80d8c0fced75","arxiv_id":"2509.03368","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"Violation of gauge anomaly cancellation conditions is equivalent, in on-shell amplitudes, to a breakdown of collinear factorization in one-loop five-point amplitudes with graviton exchange.","lead":"This paper argues that gauge anomalies, not just in Feynman diagrams, show up as a failure of collinear factorization in five-particle scattering amplitudes built from on-shell data. It uses the graviton as a universal probe to identify abelian, non-abelian, and mixed U(1)-gravitational anomalies.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The central no-go relies on isolating M3γ2ψ+R' from the rest of the 5-point amplitude; the full amplitude is not computed, so cancellation of the 1/sqrt(s12) singularity by other one-loop terms is not excluded.","rationale":"The central claim requires that no other one-loop contribution and no admissible rational completion can restore collinear factorization in the 5-point amplitude unless the anomaly coefficient vanishes. The manuscript computes only the s45-exchange contribution and then argues isolation; the self-flagged statement that other terms can be numerically similar order for s45→0 makes this the weakest point. The phase and charge arguments reduce but do not eliminate the possibility of cancellations: other diagrams with the same graviton pole in the 4-5 channel can carry the same sum_n s_n Q_n^3 factor and the same <45>/[45] phase. Similarly, the displayed R3γh in Eq. (4.4) is one possible rational completion; without a classification of all rational terms compatible with locality and 1↔2 symmetry, the claim that pathologies in other collinear limits are inevitable is not a proof. Therefore the necessity of Eq. (4.6) is conditional on the full one-loop amplitude. A direct computation of the complete 5-point amplitude in a toy model, or at least of all s45-channel unitarity cuts, would settle whether cancellation occurs. The rest of the paper, including the 4-point unitarity computations and the non-anomalous check in Appendix B, is consistent and gives real support; this is why the verdict should remain CONDITIONAL rather than REJECT or UNVERDICTED.","tokens_in":13115,"tokens_out":8493,"duration_ms":71711,"concrete_test":"Compute the complete one-loop five-point amplitude M^(1)[1γ^-2γ^-3γ^+4ψ^-5ψbar^+] for a toy theory with one charged left-handed fermion (Q=1) and one dark fermion (Q_k=0), using generalized unitarity or a direct Feynman-diagram evaluation in D=4−2ε. Extract the full coefficient of the s45-pole and the O(s45^0) terms as s45→0, then take the p1∥p2 limit. If the 1/sqrt(s12) coefficient from the complete amplitude is nonzero and cannot be removed by any rational completion, the isolation assumption is validated; if additional terms with the same <45>/[45] phase and sum_n s_n Q_n^3 dependence cancel it, Eqs. (4.3)–(4.6) need revision. A cheaper first check: compute all unitarity cuts in the s45 channel and verify the sum of residues is exactly the single term used in Eq. (4.1), rather than a sum over multiple exchange states.","verdict_should_be":"CONDITIONAL","load_bearing_attack":"The argument's linchpin is the assertion in Section 4 (after Eq. (4.2)) that M3γ2ψ+R' can be studied in isolation, so no other contribution to M^(1)[1γ^-2γ^-3γ^+4ψ^-5ψbar^+] can cancel the 1/sqrt(s12) singularity shown in Eq. (4.3). The authors themselves note that for s45→0 this term is only O(s45^0), so other terms in the amplitude can be numerically of similar order. The two supporting arguments do not close this gap. The phase argument (rotating p5 around p4 changes <45>/[45]) only distinguishes terms with a 1/s45 graviton pole in the 4-5 channel from terms without that pole; it says nothing about other one-loop diagrams that have the same s45 pole from graviton exchange but with the graviton attached to an internal charged-fermion loop rather than to the external pair. The charge-dependence argument (Q_k=0 dark fermion) excludes diagrams where the external fermion pair couples to photons, but not diagrams where the external pair couples gravitationally to a loop of charged fermions; those can carry the same sum_n s_n Q_n^3 coefficient and the same <45>/[45] phase. Moreover, R3γh in Eq. (4.4) is presented as 'a possible choice'; no exhaustive classification of rational completions is given, so the claim that no adjustment can repair the 1∥2 limit without breaking 1∥3 is not proven. Thus the necessity of Eq. (4.6) is established only for the isolated subamplitude, not for the full amplitude.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper proposes an on-shell manifestation of gauge anomalies: in a chiral gauge theory coupled to gravity, certain one-loop amplitudes with external photons/gluons and a virtual graviton exchange are proportional to anomaly coefficients, and when these are used as seeds for five-point amplitudes, the resulting collinear singularities are incompatible with the one-loop collinear factorization theorem unless the anomaly cancellation conditions hold. Section 2 contains explicit unitarity-based computations of four-point amplitudes M(1)[3γ h], M(1)[3h γ], and M(1)[3g h], proportional to the U(1)^3, mixed U(1)-gravitational, and non-abelian triangle anomaly coefficients. Section 3 reviews the factorization theorem and explains why four-point amplitudes do not expose anomalies. Section 4 constructs five-point amplitudes with a near-collinear external fermion pair and studies their 1∥2 collinear limit, finding a 1/√s12 singularity that cannot be represented by universal splitting functions; attempts to cancel it by adjusting rational terms introduce a pathological 1/s13 singularity in another collinear limit. The conclusion is that anomaly cancellation conditions (Eq. (5.1)) are necessary for collinear factorization.","tokens_in":13338,"tokens_out":8346,"duration_ms":73028,"significance":"The paper offers a clean, systematic S-matrix criterion for anomaly cancellation, with gravity as a universal probe for all three types of triangle anomalies. The four-point unitarity computations are explicit, the collinear scaling analysis is clear, and the n≥5 setup correctly avoids the kinematic degeneracy that weakens four-point arguments. If the five-point argument is made rigorous, this would be a valuable contribution to the on-shell programme. However, the central five-point claim currently rests on an isolation assumption that is asserted but not proven.","major_comments":[{"comment":"The statement that M3γ2ψ + R' can be studied 'in the sense that there cannot be cancellation' is load-bearing and not established. The paper itself notes that other terms in the amplitude can be of the same order for small s45. The phase argument by itself only shows that terms with different rotational phase cannot cancel the ⟨45⟩/[45] factor, and the charge argument with a dark fermion only rules out contributions in which the external pair couples to photons; it does not exclude other s45-singular contributions with the same phase and the same ∑_n s_n Q_n^3 coefficient. Without computing the full one-loop five-point amplitude (or at least the complete set of terms singular in both s45 and s12), the possibility that Eq. (4.3)'s 1/√s12 singularity is canceled by other terms is not excluded. Since Eq. (4.6) is the main result of the paper, this gap must be closed.","section":"Section 4, after Eq. (4.2)"},{"comment":"The rational term R3γh is introduced as 'a possible choice,' and the conclusion that no adjustment can repair the 1∥2 limit without breaking the 1∥3 limit is asserted but not proven by an exhaustive classification. Because the argument requires showing that every allowed rational completion (consistent with 1↔2 Bose symmetry and other constraints) either leaves the 1/√s12 singularity or produces a non-universal 1/s13 singularity, the paper needs either a rigorous proof or an explicit statement of the assumptions under which this classification is complete.","section":"Section 4, Eq. (4.4)"},{"comment":"The non-abelian and mixed-anomaly generalizations inherit the same isolation assumption without additional checks. For the non-abelian amplitude in Eq. (4.7), the case with two external gravitons may involve additional subtleties (e.g., graviton helicity structure and the form of the rational part), and the same 'cannot be amended' claim appears without a systematic analysis. The reader is left without a demonstration that the chosen contributions are the only ones able to develop the pathological collinear limits.","section":"Sections 4.2 and 4.3"}],"minor_comments":[{"comment":"The symbols M3γh and M3hγ are easily confused; consider renaming the cut-constructible parts, e.g., M^{cut}_{3γh} and M^{cut}_{3hγ}.","section":"Eqs. (2.7), (2.8)"},{"comment":"The expression 'M(0)[1−γ 2−γ 3+γ 4+h ] = 0' appears to be a typo; the vanishing object should be the tree-level three-photon amplitude.","section":"Text after Eq. (4.1)"},{"comment":"Reference [12] is cited as 'unpublished' without a year; please provide an updated reference or a more complete citation.","section":"Reference [12]"},{"comment":"The notation nR+ = nR_R + nR_L and nR− = nR_R − nR_L would benefit from a clarifying sentence, since the superscript R is used both for the representation and for the right/left fermion counts.","section":"Eq. (2.9)"},{"comment":"The flowchart is helpful; a brief explanation of the 'Fix R' step in the caption would improve reproducibility.","section":"Figure 2"}],"recommendation":"major_revision","confidential_remarks":"The paper is likely to be of interest to the readership given recent activity on on-shell anomaly criteria. The main risk is the isolation argument in Section 4; if the authors can supply a full five-point unitarity computation (even for a single helicity configuration with a dark fermion) or a rigorous no-cancellation proof, the paper would be strong. I would not reject on the current evidence, but the central claim cannot be accepted without addressing this point."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Ben,\n\nThe paper's claim is that gauge anomalies show up as a breakdown of one-loop collinear factorization in selected 5-point amplitudes with graviton exchange, and that the only way out is the standard anomaly cancellation conditions. This is a real step beyond earlier on-shell anomaly papers, which stayed at 4 points where kinematics is too constrained to test factorization. The graviton as a universal probe for abelian, non-abelian, and mixed anomalies is a nice unifying idea, and the unitarity bootstrap of the 4-point seed amplitudes in Section 2 is explicit and checkable. I verified the key collinear scaling in Eq. (4.3): the cut-constructible part behaves as 1/sqrt(s12), which no universal splitting function can produce. The non-anomalous sanity check in Appendix B is a good discipline.\n\nThe soft spots are real but not disqualifying. The paper does not compute the full 5-point amplitude; it isolates the s45 graviton-exchange contribution and argues that no other term can cancel its pathological 1/sqrt(s12) singularity. The phase and charge arguments are plausible, and the dark-fermion limit (Q_k=0) makes the isolation airtight, but the general case is asserted more than proven. The larger gap is rational-term uniqueness: the paper exhibits one rational completion R3γh that fixes the 1||2 limit but breaks the 1||3 limit, then states without proof that no adjustment can repair both. I would want either a full 5-point computation or a classification of the allowed rational terms before calling the no-go a theorem. The citation pattern is appropriate; the only wart is reliance on unpublished Ref. [12] for the 4-point non-violation.\n\nWho is this for? On-shell amplitude practitioners and anyone working on consistency conditions for chiral gauge theories. The central claim is likely true, but the proof is incomplete. I would send it to a serious referee. If the isolation and rational-term gaps can be closed, it becomes an important result; as it stands, it is a strong, well-argued suggestion.","headline":"A credible on-shell diagnosis of gauge anomalies via collinear factorization breakdown, with the main no-go resting on an unproven rational-term uniqueness claim.","tokens_in":13968,"tokens_out":8586,"would_cite":true,"duration_ms":79298,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":["81T50","81T18"],"pacs":[],"model":"deepseek-v4-flash","headline":"The paper argues that gauge anomalies manifest on shell as a breakdown of one-loop collinear factorization in five-point amplitudes, and that the only cure is the standard anomaly cancellation conditions.","keywords":["gauge anomalies","collinear factorization","on-shell amplitudes","unitarity method","one-loop amplitudes","graviton exchange","chiral fermions","anomaly cancellation"],"falsifier":"Compute the complete one-loop five-point amplitude $M^{(1)}[1^-_\\gamma 2^-_\\gamma 3^+_\\gamma 4^-_\\psi 5^+_{\\bar\\psi}]$ in an anomalous theory, keeping every channel and every rational term; if the $1/\\sqrt{s_{12}}$ singularity of the isolated graviton-exchange piece cancels against the remaining terms, collinear factorization could survive even with $\\sum_n s_n Q_n^3\\neq 0$, and the central claim would be wrong.","tokens_in":12794,"feed_emoji":"⚛️","tokens_out":12146,"duration_ms":98171,"temperature":0.7,"pith_summary":"This paper argues that a gauge theory with chiral fermions that fails anomaly cancellation cannot satisfy universal one-loop collinear factorization. The authors construct five-point amplitudes built from a one-loop four-point amplitude with three photons and a graviton, sewn by unitarity onto a fermion pair, and show that in an anomalous theory the cut-constructible part develops $1/\\sqrt{s_{12}}$ collinear singularities that no universal splitting function can reproduce. Adjusting the rational term to remove that singularity creates new pathologies in other collinear limits, so the only resolution is to impose the standard cancellation conditions $\\sum_n s_n Q_n^3=0$, $\\sum_n s_n Q_n=0$, and $\\sum_R n_R^- d_R^{abc}=0$. The argument never invokes a Lagrangian or Ward identity, so it proposes anomaly cancellation as a direct consequence of unitarity plus collinear factorization.","feed_headline":"Gauge anomalies break collinear factorization","feed_subtitle":"Without Lagrangians or Ward identities, collinear factorization reproduces the standard anomaly cancellation conditions.","key_machinery":"The load-bearing object is the isolated graviton-exchange contribution to the five-point amplitude, written as $M_{3\\gamma 2\\psi}+R'_{3\\gamma 2\\psi}$ in Eq. (4.2), obtained by sewing the one-loop four-point amplitude $M^{(1)}[1^-_\\gamma 2^-_\\gamma 3^+_\\gamma 4^+_h]$—proportional to $\\sum_n s_n Q_n^3$—onto the tree-level graviton-fermion amplitude. The identity that makes the argument work is that this piece is separable from the rest of the amplitude: the factor $\\langle 45\\rangle/[45]$ picks up a large phase when one fermion momentum is rotated around the other, and the charge dependence is shared by no other term that could develop the collinear singularity. What carries the proof is the one-loop collinear factorization theorem stated in Refs. [13, 14], which says that for $n\\ge 5$ every collinear limit must factor into a universal splitting function times a lower-point amplitude; the pathological $1/\\sqrt{s_{12}}$ behavior fits none of the allowed photon or graviton splittings.","core_discovery":"On the authors' own terms, the central claim is that gauge anomalies appear in the on-shell S-matrix as a violation of collinear factorization: in an anomalous chiral theory, selected one-loop five-point amplitudes have a $1/\\sqrt{s_{12}}$ singularity in the $1\\parallel 2$ collinear limit that cannot be written as a universal splitting function times a lower-point amplitude. The offending singularity comes from the graviton-exchange channel, where the anomaly coefficient of the one-loop four-point amplitude enters the residue. Restoring factorization by adding a rational term is never a fix: the term needed to cure $1\\parallel 2$ produces a $1/s_{13}$ singularity in the $1\\parallel 3$ limit, which is equally incompatible with the theorem and cannot be removed without breaking Bose symmetry. The non-abelian triangle anomaly and the mixed $U(1)$-gravitational anomaly obey the same pattern, so all three standard cancellation conditions follow from collinear-factorization consistency alone.","pith_inferences":["One could turn the phase isolation argument into a practical test: rotate one external fermion momentum around its partner and look for terms with the $\\langle 45\\rangle/[45]$ phase; any such term not reproduced by a splitting function is an anomaly signal.","The same construction should apply to axion-like exchange mechanisms: adding an axion-like state in the graviton channel would contribute to the same residue and could restore collinear factorization, giving an on-shell analogue of that mechanism.","For anomaly-free but chiral theories with anomalous global symmetries, gravity-probed five-point amplitudes might expose physical signatures of the global anomaly rather than an inconsistency, since factorization need not be saved in that case.","Because the singularities require massless chiral fermions, the on-shell probe is specifically sensitive to the massless chiral content of the theory."],"forward_implications":["An anomalous chiral gauge theory cannot be embedded in a unitary, local on-shell theory, because its five-point amplitudes violate the one-loop collinear factorization theorem.","The three standard anomaly cancellation conditions emerge as the unique on-shell consistency conditions: $U(1)^3$, mixed $U(1)$-gravitational, and non-abelian triangle.","Four-point amplitudes are not enough to expose the pathology, so future on-shell anomaly tests should target five-point collinear limits with a graviton attached.","Gravity acts as a universal probe: coupling a single graviton to a photonic or gluonic loop converts the anomaly coefficient into a testable singular behavior.","Because adjusting rational terms only moves the problem from one collinear channel to another, no local rational counterterm can repair an anomalous theory."],"supporting_citations":[{"why":"Supplies the one-loop collinear factorization theorem that the five-point amplitudes are tested against.","marker":"[13]"},{"why":"Extends the factorization statement and supplies the universal one-loop splitting-function framework.","marker":"[14]"},{"why":"Earlier on-shell consistency argument for the four-gluon amplitude whose unitarity-based method this paper adapts.","marker":"[8]"},{"why":"Constructs one-loop anomalous amplitudes without an action, the technical starting point for the unitarity bootstrap here.","marker":"[9]"},{"why":"Observed that abelian anomalies appear in four-point photon-graviton amplitudes, the seed of the gravity-as-probe idea.","marker":"[11]"},{"why":"Shows that four-point u-pole residues vanish on real momenta, motivating the move to five-point kinematics.","marker":"[12]"},{"why":"Establishes that gravitational splitting functions are tree-level exact, ruling out a graviton splitting as an explanation of the square-root singularity.","marker":"[20]"},{"why":"Furry's theorem guarantees the anomaly-proportional four-point amplitude vanishes in non-chiral theories, isolating the anomaly coefficient.","marker":"[16]"}],"fun_headline_variants":["Anomalies break collinear factorization on-shell","Graviton exchange reveals anomalies in collinear limits","Factorization fails precisely where gauge anomalies live","Collinear singularities expose gauge anomaly conditions","On-shell proof: anomalies violate collinear factorization"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"Everything rests on the assumption that the graviton-exchange piece isolated in Eq. (4.2) can be studied by itself, so no other term in the complete one-loop five-point amplitude cancels its bad collinear singularities.","fun_headline_variants_meta":{"raw":{"variants":["Anomalies break collinear factorization on-shell","Graviton exchange reveals anomalies in collinear limits","Factorization fails precisely where gauge anomalies live","Collinear singularities expose gauge anomaly conditions","On-shell proof: anomalies violate collinear factorization"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.001051,"raw_usage":{"total_tokens":4331,"prompt_tokens":782,"completion_tokens":3549,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":398,"completion_tokens_details":{"reasoning_tokens":3480}},"tokens_in":398,"tokens_out":3549,"duration_ms":22814,"temperature":1.0,"reasoning_tokens":3480,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-15T16:31:39.297890+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Compute the complete one-loop five-point amplitude $M^{(1)}[1^-_\\gamma 2^-_\\gamma 3^+_\\gamma 4^-_\\psi 5^+_{\\bar\\psi}]$ in an anomalous theory, keeping every channel and every rational term; if the $1/\\sqrt{s_{12}}$ singularity of the isolated graviton-exchange piece cancels against the remaining terms, collinear factorization could survive even with $\\sum_n s_n Q_n^3\\neq 0$, and the central claim would be wrong.","supporting_citations":[{"cited_title":"Bonnefoy, S","cited_arxiv_id":null,"evidence_quote":"Shows that four-point u-pole residues vanish on real momenta, motivating the move to five-point kinematics."},{"cited_title":"Furry,A Symmetry Theorem in the Positron Theory, Phys","cited_arxiv_id":null,"evidence_quote":"Furry's theorem guarantees the anomaly-proportional four-point amplitude vanishes in non-chiral theories, isolating the anomaly coefficient."}],"review_version":2}