{"id":"3e11c3fb-9f35-4e34-9a40-7e91fa22a721","arxiv_id":"2412.09344","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":3,"one_line_summary":"Existing LHC data exclude fat-brane mUED compactification scales below about 3 TeV, and ML-tagged searches at 500 fb^-1 could extend the reach to roughly 3.2 to 3.4 TeV.","lead":"This paper updates the collider bounds on the fat-brane version of the minimal universal extra dimension model using the latest ATLAS photon, di-photon, and multijet searches, then proposes boosted-decision-tree search strategies for future LHC runs. If the projections hold, the new strategies could push the model's compactification-scale reach to about 3.2 to 3.4 TeV at 500 inverse femtobarns.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"GMD width calculation is the linchpin; the density-of-states integral (Eq. 3.2) is not derived for the fat-brane tower and Eq. A.2 contains an apparent typo, so all quoted bounds could shift.","rationale":"The recasting methodology itself is well supported: the mono-photon cut-flow tables agree with ATLAS HepData to within ~10–20% at each stage, the multijet recasting was validated previously in Ref. [18], and the diphoton bound reproduces Ref. [18]. This is real evidence for the experimental-side implementation. What is not independently verified is the model-side GMD width, which enters every acceptance calculation. The reader identified this as the weakest assumption; I agree. The concern is not that the authors are wrong, but that the central numerical results inherit a calculation that is (a) not benchmarked against any direct summation, and (b) presented in the appendix with an apparent typo that blocks reproduction. The honest verdict remains CONDITIONAL: the bounds are plausible but should not be treated as definitive until the width calculation is checked. No other issue (BDT truth-tagging optimism, diphoton extrapolation, 20% background uncertainty) is as load-bearing, since those affect only the projections, whereas the GMD width affects the current exclusions as well.","tokens_in":41194,"tokens_out":11541,"duration_ms":117816,"concrete_test":"Recompute the total GMD widths of g1 (g1→g+G) and γ1 (γ1→γ+G) at a benchmark point (e.g., R^{-1}=3.1 TeV, M_D=5 TeV, N=2 and N=6) by explicitly summing over the KK modes n5=0..N_5 and the large-dimension lattice up to |n| such that m_n ≤ M, using the corrected form factors from Refs. [94,96] (and the intended version of Eq. A.2). Compare with the one-dimensional integral Eq. (3.2). If the ratio deviates from unity by more than ~10%, the integral approximation and all derived bounds need revision. As a second check, recompute the N=6 diphoton bound of Ref. [18] with the corrected formula to see whether the claimed validation still holds.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The quantitative backbone of every exclusion and projection is the gravity-mediated decay (GMD) width. The paper inherits the per-mode widths from Refs. [94,96] and replaces the sum over the graviton KK tower by Eq. (3.2): Γ = (M_Pl^2/M_D^{N+2}) ∫ dΩ dm m^{N-1} Γ_n. The exponent N-1 is the density of states of the N large extra dimensions. However, the graviton tower in this fat-brane model includes the small (TeV^{-1}) universal extra dimension as well; the form factors in Eq. (A.2) depend explicitly on n5, the KK index along that dimension. No derivation is given showing that the sum over n5 and the N large dimensions collapses to this simple m^{N-1} integral. For N=6, where the spectrum peaks at high graviton masses (Fig. 1), threshold and Jacobian corrections from the n5-dependent mass shell could be O(1). In addition, the printed Eq. (A.2) states |F^s|^2 = |F^s|^2/x5^2, an identity that is self-referential and cannot be implemented; presumably |F^s|^2 = |F^c|^2/x5^2 was intended. Because the paper validates its GMD implementation only against Ref. [18] (which uses the same untested formulas) and against diphoton kinematics, an error here would shift all bounds in Fig. 5 and projections in Fig. 15 by an amount proportional to the width error.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies the collider phenomenology of the fat-brane realization of minimal Universal Extra Dimensions (mUED), in which gravity-mediated decays (GMD) of level-1 KK particles produce final states with photons, jets, massive bosons, and missing transverse energy. It recasts ATLAS mono-photon (139 fb^-1), multijet (139 fb^-1), and diphoton (36.1 fb^-1, extrapolated to 139 fb^-1) searches to derive 95% CL exclusion limits on the R^-1--MD plane for N = 2, 4, 6, and it proposes BDT-based search strategies with fat-jet tagging for 500 fb^-1 of future LHC data. The main quantitative claims are that the N = 2 multijet search excludes R^-1 below about 2975 GeV, the N = 6 diphoton reach reaches about 3000 GeV at MD = 15 TeV, and the proposed optimized analyses extend the reach to about 3320 GeV (N = 2), 3190 GeV (N = 4), and 3390 GeV (N = 6) at 500 fb^-1.","tokens_in":41627,"tokens_out":12698,"duration_ms":129866,"significance":"If the GMD width calculation is correct, the paper provides a useful and reasonably complete update of the LHC constraints on this specific extra-dimensional scenario. It has genuine strengths: the mono-photon recasting is validated against ATLAS HEPData cut-flow tables (Tables 3, 8, 9), the bounds are anchored to official 95% CL visible cross-section limits, and the object and signal-region definitions are documented in enough detail for the analysis to be reproduced. The future projections with fat-jet tagging and BDT classifiers are a constructive addition. However, because every quoted exclusion and projection is proportional to the GMD width, the lack of a derivation for the density-of-states integral and the apparent typo in the form-factor definition are load-bearing issues that must be addressed before the numerical results can be fully trusted.","major_comments":[{"comment":"The replacement of the graviton KK sum by the integral Gamma = (M_Pl^2/M_D^{N+2}) integral dm dOmega Gamma_n m^{N-1} assumes a density of states m^{N-1} for the tower of large extra dimensions. In the fat-brane setup described in Section 2.1, gravity also propagates in the TeV^-1 universal dimension, and the form factors in Eq. (A.2) depend explicitly on n5. No derivation is given showing that the sum over n5 and the N large dimensions collapses to this simple integral, and for N = 6, where Fig. 1 shows the spectrum peaking at large graviton masses, threshold and Jacobian effects could alter the result by O(1). Since every quoted bound and projection is proportional to this width, please provide the derivation or justify the approximation explicitly.","section":"Section 3.2, Eq. (3.2)"},{"comment":"The printed relation |F^s_{l|n}|^2 = |F^s_{l|n}|^2 / x5^2 is self-referential and cannot be implemented as written; presumably |F^s|^2 = |F^c|^2/x5^2 or an equivalent corrected expression was intended. Please state the exact form factors used in the modified PYWIDTH routine and confirm that the numerical results are stable under the correction.","section":"Appendix A.1, Eq. (A.2)"},{"comment":"The 139 fb^-1 diphoton limits are not ATLAS results but an extrapolation of the 36.1 fb^-1 analysis assuming an overall 40% background uncertainty. This extrapolated limit is used in Fig. 5 and is the strongest constraint for N = 6 at large MD (R^-1 around 3000 GeV). The paper should present the sensitivity of this bound to the assumed uncertainty and clearly label the diphoton contours as expected limits rather than observed constraints.","section":"Section 3.4.3 and Table 4"},{"comment":"The BDT V-tagger is trained on WW/ZZ and dijet samples and then applied to signal fat jets from KK cascade decays without demonstrating that the tagging efficiency transfers. Because the V-tagged jet multiplicity is a key input to the projections in Figs. 12-15, please report the signal-side tagging rate as a function of R^-1 and MD, or provide a mixed-signal validation, and quantify the resulting systematic uncertainty.","section":"Sections 3.5.1 and Appendix A.3"},{"comment":"The multijet recast is not validated with a cut-flow table in this work but is inherited from Ref. [18], which uses the same GMD implementation. The N = 2 exclusion (R^-1 below 2975 GeV) rests entirely on this channel, so a cross-check of the multijet signal-region efficiencies against an external implementation, or at least a cut-flow table analogous to Tables 3, 8 and 9, is needed to support the central bound.","section":"Section 3.4.2"}],"minor_comments":[{"comment":"The abstract contains the typo \"139 inverse femtobern\"; it should read \"139 inverse femtobarns\".","section":"Abstract"},{"comment":"The caption labels the upper right panel as \"N = 2\", but from the text and the surrounding panels it should be \"N = 4\".","section":"Fig. 15 caption"},{"comment":"The KK decomposition of the graviton uses the index n5 for the small dimension but writes the exponential and the mass formula with the large-dimension radius r; the distinction between R and r should be clarified to avoid ambiguity about which radius controls the n5 mode mass.","section":"Section 2.1, Eqs. (2.1) and (2.7)"},{"comment":"The signal region is referred to as SR_0pnl in the table header but as SR_0p0l and SR_0pnj elsewhere; please make the naming consistent.","section":"Table 7 and Section 3.5.2"},{"comment":"The invariant mass formula is written as M = sqrt(sum_i E_i^2 - sum_i p_i^2); the correct jet invariant mass is sqrt((sum_i E_i)^2 - (sum_i p_i)^2). Please correct this.","section":"Appendix A.3, Eq. (A.4)"},{"comment":"The numbers quoted for the N = 4 future reach are inconsistent: Section 3.5.3 states 3190 GeV at MD around 14 TeV, while Section 4 reports 2942(3157) GeV for MD = 5000(15000) GeV; please align the text and figures.","section":"Section 3.5.3 and Section 4"}],"recommendation":"major_revision","confidential_remarks":"The central risk is that the GMD width implementation, which underpins all numerical results, is inherited from previous papers and the manuscript does not resolve the density-of-states question or the form-factor typo. The cut-flow validation against ATLAS is a genuine strength, but it mostly validates the detector simulation and object selection, not the GMD width itself. I would advise the editor that the paper is appropriate in scope for JHEP, provided the authors either supply the missing derivation or explicitly demonstrate that the quoted bounds are insensitive to the n5 treatment and to the corrected form factors."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Useful update on fat-brane mUED. The real news is the recast of the 139/fb ATLAS mono-photon search and the new BDT-tagged boosted-boson signal regions; the authors reproduce ATLAS cut-flows well, and the projected 500/fb reach is a concrete, testable proposal. The bounds on R^-1 around 3 TeV are plausible if the gravity-mediated width implementation is right.\n\nThat 'if' is where I'd focus. Eq. (3.2) replaces the sum over the whole KK graviton tower by an integral with density of states m^{N-1}, but the tower also includes the small extra dimension and the form factors in Eq. (A.2) depend on n5. I don't see a derivation that the n5 sum collapses to that simple integral, and for N=6 the spectrum peaks at large graviton mass, so threshold/Jacobian corrections could be O(1). On top of that, Eq. (A.2) as printed is self-referential (|F^s|^2 = |F^s|^2/x5^2); presumably a typo for |F^c|^2/x5^2, but it shows the appendix wasn't proofread. The authors validate against the diphoton analysis of Ref [18], which uses the same formulas, so that's not independent. A mistake here would shift every bound in Fig. 5 and projection in Fig. 15.\n\nOther soft spots are minor. The diphoton projection to 139/fb uses an assumed 40% background uncertainty, and the future projections assume 20%; both are stated and standard for this kind of study, but they deserve a sensitivity check. No code or data released, which for a recast paper is a bit unfortunate.\n\nIf the GMD width issue is resolved—ideally by a derivation or a cross-check against an independent calculation—the paper is a useful contribution. As it stands, the bounds are conditional. I'd send it to a good referee who knows extra-dimensional models as well as recasts, and ask for that clarification plus a few plots showing how much the limits move under reasonable variations of the width formula.","headline":"A solid, well-validated recast of LHC searches onto fat-brane mUED with new BDT-tagged projections, but the gravity-mediated width calculation is the linchpin and needs scrutiny before the bounds are taken at face value.","tokens_in":42085,"tokens_out":2237,"would_cite":false,"duration_ms":22812,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"Recast LHC searches already push the fat-brane mUED scale above about 3 TeV, and the paper projects that BDT-based searches could reach about 3.4 TeV.","keywords":["universal extra dimensions","fat brane","gravity-mediated decays","Kaluza-Klein graviton tower","LHC recasting","mono-photon plus MET","boosted boson tagging","missing transverse energy"],"falsifier":"Recompute the level-1 KK-gluon gravity-mediated width by explicitly summing over the graviton tower for $N=2,4,6$ at, say, $R^{-1}=3$ TeV and $M_D=5$ TeV, and compare with the density-of-states integral of Eq.~(3.2); a deviation larger than the Monte-Carlo statistical uncertainty would change the branching ratios and therefore the recast bounds such as $R^{-1}<2975$ GeV for $N=2$. A dedicated search with the proposed five signal regions at $500\\ \\mathrm{fb}^{-1}$ would also settle whether the predicted event rates appear.","tokens_in":40990,"feed_emoji":"⚛️","tokens_out":14494,"duration_ms":122955,"temperature":0.7,"pith_summary":"The paper targets the fat-brane version of the minimal Universal Extra Dimension model (FB-mUED), in which Standard Model fields live on a TeV-sized fifth dimension while gravity also fills additional large dimensions. Because each Kaluza-Klein (KK) excitation can decay into a whole tower of graviton modes, KK particles produce hard photons, jets, and boosted $W/Z/H$ bosons or top quarks together with large missing transverse energy. The paper recasts three 13 TeV LHC searches for these final states and finds that the multijet channel alone excludes $R^{-1}$ below about 2975 GeV for $N=2$ large extra dimensions, with complementary mono-photon and di-photon bounds for $N=4$ and $N=6$. It then designs five BDT-based signal regions that tag boosted bosons and projects that $500\\ \\mathrm{fb}^{-1}$ would extend the 95% C.L. reach to roughly 3320 GeV ($N=2$) and up to about 3390 GeV ($N=6$). If these projections hold, the model is already pressed to multi-TeV radii and future LHC runs could probe even higher scales.","feed_headline":"New LHC bounds push extra-dimension KK masses past 3 TeV","feed_subtitle":"Recasting LHC photon and jet searches gives the tightest fat-brane mUED bounds yet; BDT searches could reach ~3.4 TeV.","key_machinery":"The load-bearing mechanism is the gravity-mediated decay width, computed by replacing the sum over the tower of Kaluza-Klein graviton modes by an integral over the graviton density of states, $\\Gamma = (M_{Pl}^2/M_D^{N+2}) \\int dm\\, d\\Omega\\, \\Gamma_{\\vec n}\\, m_{\\vec n}^{N-1}$ (Eq.~3.2). This replacement makes the otherwise Planck-suppressed single-mode widths sizable, because each KK state can decay into many kinematically allowed graviton modes, and it sets the relative strength of GMD versus KK-number-conserving decays that fixes every final state. For the future projections, the second central object is a boosted decision tree (BDT) classifier trained on fat-jet substructure variables (invariant mass, N-subjettiness, jet charge, b-tag) to tag $W/Z$ and top jets against QCD jets, which suppresses the dominant backgrounds in the five proposed signal regions.","core_discovery":"The paper establishes that the fat-brane mUED parameter space is strongly constrained by existing LHC searches originally designed for supersymmetry, once those searches are recast with gravity-mediated decays. The key finding is scenario-dependent: for $N=2$ the large graviton density of states makes gravity-mediated decay dominate, so the multi-jet plus missing-energy search gives the best bound, excluding $R^{-1}<2975$ GeV independently of $M_D$; for $N=4$ the multijet and mono-photon searches are complementary, excluding $R^{-1}$ below about 2898 GeV ($M_D=5$ TeV) and 2958 GeV ($M_D=15$ TeV); for $N=6$, where KK-number-conserving cascades dominate, the di-photon channel is most sensitive, with an expected reach near 3000 GeV at $139\\ \\mathrm{fb}^{-1}$. The paper further claims that its own five signal regions with BDT-based boosted-boson tagging would, at $500\\ \\mathrm{fb}^{-1}$, improve the $N=2$ reach to about 3320 GeV and reach about 3390 GeV for $N=6$ at high $M_D$.","pith_inferences":["The same recasting method could be applied to the corresponding search channels from the other large LHC detector; using both detectors' data would tighten or independently cross-check these bounds.","Because the graviton tower makes the photon $p_T$ spectrum softer than in gauge-mediated supersymmetry, future experimental signal regions may do better with lower photon $p_T$ thresholds combined with higher missing-energy requirements than with the SUSY-oriented cuts.","Before quoting the $N=6$ projection, the discrepancy between the roughly 3390 GeV reach stated in Section 3.5.3 and the roughly 3225 GeV value in Section 4 should be resolved; the final limit line could move by several hundred GeV.","The BDT taggers could be validated on Standard Model $W/Z$+jets data at 13 TeV before being used in a new physics search, which would test tagger performance independently of the model."],"forward_implications":["Existing LHC photon-plus-MET, di-photon-plus-MET, and multijet-plus-MET searches, recast model-independently, already exclude fat-brane mUED level-1 KK gluon and photon masses below roughly 3 TeV for all $N=2,4,6$.","For $N=2$, the multijet channel dominates and further luminosity mainly extends the jet-based reach; the reach is nearly independent of $M_D$.","For $N=4$ and $N=6$, the mono-photon and di-photon channels become the most sensitive at high $M_D$, so the constraints are genuinely two-dimensional in the $R^{-1}$--$M_D$ plane.","The proposed BDT signal regions, combined statistically, project $500\\ \\mathrm{fb}^{-1}$ 95% C.L. reaches of about 3320 GeV for $N=2$ and up to about 3390 GeV for $N=6$, roughly 345 GeV beyond the current $N=2$ bound.","Boosted boson tagging and lepton-inclusive signal regions recover signal that SUSY-tailored cuts reject, so the model's reach is not limited by the existing search optimizations."],"supporting_citations":[{"why":"Supplies the mono-photon plus missing-transverse-energy search's signal regions (SRL, SRM, SRH) and their model-independent 95% CL visible cross-section limits used to constrain the $R^{-1}$--$M_D$ plane.","marker":"[1]"},{"why":"Supplies the multijet-plus-MET search's ten signal regions and 95% CL cross-section limits; this recast produces the strongest $N=2$ bound.","marker":"[2]"},{"why":"Supplies the di-photon-plus-MET search's signal regions and 36.1 fb$^{-1}$ limits, which the paper extrapolates to 139 fb$^{-1}$.","marker":"[3]"},{"why":"Prior recast of UED models with gravity-mediated decays; provides the diphoton and multijet implementation the paper validates against.","marker":"[18]"},{"why":"Gives the gravity-matter Feynman rules and the gravity-mediated decay width formulas for KK fermions and gauge bosons.","marker":"[96]"},{"why":"Introduces the density-of-states integral for the graviton tower that turns individual suppressed widths into a sizable total GMD width.","marker":"[100]"},{"why":"Provides the radiative corrections that split the level-1 KK masses and set the KK-number-conserving decay pattern.","marker":"[77]"},{"why":"Establishes the consistency of minimal and non-minimal UED with 13 TeV LHC data and supplies the multijet recast setup.","marker":"[47]"}],"fun_headline_variants":["LHC recast pushes extra-dimension KK masses past 3 TeV","Gravity-mediated decays yield strongest fat-brane mUED bounds","Machine learning extends extra-dimension search to 3.4 TeV","New LHC searches constrain fat-brane extra dimensions","Excluded extra-dimension masses climb past 3 TeV at LHC"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"Every bound and projection inherits the gravity-mediated decay widths, computed by replacing the sum over the graviton tower with an integral over the graviton density of states (Eq.~3.2) and implemented in a modified shower routine; the paper validates the di-photon implementation against an earlier recast but does not separately verify the mono-photon and multijet implementations, so an error there would shift all quoted limits and reaches.","fun_headline_variants_meta":{"raw":{"variants":["LHC recast pushes extra-dimension KK masses past 3 TeV","Gravity-mediated decays yield strongest fat-brane mUED bounds","Machine learning extends extra-dimension search to 3.4 TeV","New LHC searches constrain fat-brane extra dimensions","Excluded extra-dimension masses climb past 3 TeV at LHC"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000234,"raw_usage":{"total_tokens":1522,"prompt_tokens":994,"completion_tokens":528,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":610,"completion_tokens_details":{"reasoning_tokens":435}},"tokens_in":610,"tokens_out":528,"duration_ms":5713,"temperature":1.0,"reasoning_tokens":435,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-11T17:05:48.547371+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Recompute the level-1 KK-gluon gravity-mediated width by explicitly summing over the graviton tower for $N=2,4,6$ at, say, $R^{-1}=3$ TeV and $M_D=5$ TeV, and compare with the density-of-states integral of Eq.~(3.2); a deviation larger than the Monte-Carlo statistical uncertainty would change the branching ratios and therefore the recast bounds such as $R^{-1}<2975$ GeV for $N=2$. A dedicated search with the proposed five signal regions at $500\\ \\mathrm{fb}^{-1}$ would also settle whether the predicted event rates appear.","supporting_citations":[{"cited_title":"Radiative Corrections to Kaluza-Klein Masses","cited_arxiv_id":"hep-ph/0204342","evidence_quote":"Provides the radiative corrections that split the level-1 KK masses and set the KK-number-conserving decay pattern."}],"review_version":1}