{"id":"1b9254fa-0c9a-463b-9953-2cc3cec26e6f","arxiv_id":"2608.11040","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":4,"one_line_summary":"Simulations of four scNMSSM benchmark points suggest an exotic 8b di-Higgs channel may be visible at the HL-LHC, while a boosted 8tau channel falls outside existing search strategies.","lead":"This paper simulates how often di-Higgs events in a supersymmetric model would appear as eight bottom-quark jets or eight highly collimated tau leptons at the High-Luminosity LHC. It finds that the boosted tau signature is a gap in current search coverage and argues that dedicated sub-jet taggers are needed to probe it.","discovery_kind":"new_application","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The claim that the boosted 8tau final state is 'not addressed by any existing ATLAS or CMS search strategy' is broader than the abstract's 'di-Higgs search' and is supported only by three di-Higgs references; existing h->aa->4tau and low-mass ditau resonance searches may already cover this topology.","rationale":"The reader's weakest_assumption focuses on borrowed backgrounds with non-identical selections, which is a valid concern for the 8b significance projections and BP4 comparisons. However, the reader's own strongest_claim centers on the BP2 boosted 8tau topology and its stated novelty. The single most load-bearing condition for that central claim is not the background model but the assertion that no existing ATLAS or CMS search addresses the topology. That assertion appears in the abstract in a narrow form ('di-Higgs search') and in the conclusion in a much broader form ('any existing search strategy'). The paper's citation base for this assertion is limited to three di-Higgs analyses and does not include single-Higgs exotic decay searches or low-mass ditau resonance searches, which are the most plausible existing constraints on the same a1->tautau signature. A literature and reinterpretation check is therefore both concrete and decisive: if an existing 4tau or ditau resonance search has any real acceptance for BP2, the novelty claim fails; if not, it survives. This does not change the reader's CONDITIONAL verdict because the paper is otherwise honest about its limitations and the simulation results appear internally consistent, but it sharpens the condition: the overlap with existing tau-pair searches should be checked before accepting the headline novelty. The verdict remains conditional pending that check, so UNCHANGED is the appropriate recommendation.","tokens_in":15004,"tokens_out":10022,"duration_ms":94703,"concrete_test":"Take the most recent ATLAS or CMS search for h->aa->4tau with a light a1 (e.g., CMS HIG-15-011 or a 13 TeV successor), reproduce its trigger, tau-jet, and event-selection requirements in the same Delphes 3 chain used for BP2, and compute the expected number of BP2 signal events passing those selections at 3000 fb^-1. If this number is non-negligible (conservatively, >1 event), then the conclusion that no existing ATLAS or CMS search strategy addresses the boosted 8tau topology is falsified.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper's central novelty claim is twofold: (i) for BP2, 82.4% of same-a1 tau pairs have Delta R < 0.4 and merge within a standard jet cone, and (ii) this boosted 8tau topology is 'not covered by any existing ATLAS or CMS di-Higgs search' (abstract) and 'not addressed by any existing ATLAS or CMS search strategy' (conclusion). The first part is supported by a truth-level Delta R distribution, but the second is under-supported: the paper cites only Refs. [14-16], all di-Higgs analyses, and does not survey ATLAS/CMS searches for exotic Higgs decays h->aa->4tau or for low-mass resonances decaying to tau pairs. Such searches target exactly the a1->tau+tau- decay with light a1 and would face the same merged-tau challenge; if any of them has non-negligible acceptance for BP2, the broader conclusion is overstated. The paper also does not check whether the resolved-tau selections used by existing 4tau analyses retain some efficiency for BP2 even when individual tau pairs are partially merged. Because the strongest advertised outcome is the identification of a qualitatively new channel, the literature claim should be verified before it is treated as established. This is a bounded, factual concern: the simulation numbers may be correct while the 'not covered' statement is false.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"This paper studies di-Higgs prospects in the semi-constrained NMSSM at the HL-LHC (13.6 TeV, 3000/fb) using four benchmark points from a previous parameter scan. It simulates the signal chains h1h1 -> 4a1 -> 8b (BP1, BP3), h1h1 -> 4a1 -> 8tau (BP2), and SM-like h1h1 -> bbbb / bbtautau (BP4) with MadGraph5+Pythia8+Delphes3. For BP2, with m_a1 = 8.2 GeV, it reports that 82.4% of same-a1 tau pairs have Delta R < 0.4, producing a merged, boosted topology. Signal efficiencies and yields are given, and significances are estimated using literature background yields. The central claim is that the boosted 8tau topology is not covered by existing ATLAS/CMS di-Higgs searches and requires dedicated sub-jet taggers.","tokens_in":15260,"tokens_out":12453,"duration_ms":113830,"significance":"If the benchmark points are representative of the viable scNMSSM parameter space, the paper gives a useful taxonomy of di-Higgs final states at large lambda / low tan beta. The simulation chain is standard and the reported Delta R distribution and efficiency numbers are internally consistent; the paper is also honest in Sec. 5.4 about the limitations of its background model. The identification of a boosted 8tau topology with a high merging fraction is a genuine observation that could be of interest for HL-LHC searches. However, the significance projections rest on unmatched literature backgrounds, and the assertion that the topology is 'not covered' by existing searches is broader than the citation base supports. With those issues repaired or appropriately qualified, the paper would be a solid phenomenological contribution; as it stands, the strength of some claims exceeds the evidence presented.","major_comments":[{"comment":"The abstract states that the BP2 boosted 8tau topology 'is not covered by any existing ATLAS or CMS di-Higgs search', and Sec. 6.2 states it is 'not addressed by any existing ATLAS or CMS search strategy'. The only searches cited in support are Refs. [14-16], which are standard di-Higgs analyses. The paper does not survey ATLAS/CMS searches for exotic Higgs decays h -> aa -> 4tau or for low-mass resonances decaying to tau pairs; such searches target exactly the a1 -> tau+tau- decay with a light a1 and face the same merged-tau challenge. If any of these has non-negligible acceptance for BP2, the 'not covered' claim is overstated. Because this novelty claim is a central advertised result, please either provide a systematic literature survey showing that no existing search has relevant acceptance, or qualify the claim to 'not covered by the di-Higgs searches cited here'. The discrepancy between 'di-Higgs search' in the abstract and 'any existing ... search strategy' in the conclusion should also be resolved.","section":"Abstract; Sec. 4.5; Sec. 6.2"},{"comment":"All significance projections in Table 6 use background yields taken from published ATLAS/CMS analyses whose event selections are, as Sec. 5.4 explicitly admits, not identical to the selections used in this paper; no background Monte Carlo, trigger model, or pileup treatment is included. The paper calls the estimates 'indicative', yet the abstract and Sec. 6.1 state that BP1/BP3 are 'potentially compatible with 5-sigma sensitivity' based on these numbers. Since the background yields are not derived for the present selections, the 5-sigma statements are not established. Please either simulate (or otherwise derive) backgrounds under the analysis selections used in Table 5, or clearly downgrade the 5-sigma language to a signal-to-background check with the explicit caveat that no discovery significance can be concluded. The current wording overstates the strength of the evidence.","section":"Sec. 5.2-5.4; Table 6"}],"minor_comments":[{"comment":"The sentence beginning 'However, even after applying HT > 100 GeV and >=3tau requirements, the estimated ZZ -> 4tau background alone The total background...' is incomplete and should be rewritten.","section":"Sec. 6.2"},{"comment":"The factor-of-~21 enhancement of the BP2 >=2tau efficiency over the 'SM HH baseline of ~1.8%' is quoted in the abstract, but the baseline is not defined or derived in the paper. Ref. [1] is a theory paper rather than a detector-level SM HH -> 4tau simulation; please state the exact selection that gives 1.8% or remove this comparison.","section":"Sec. 4.2"},{"comment":"Sec. 3.3 does not state whether the Delphes CMS card includes pileup; Sec. 5.4 later says pileup is not modelled, so the two sections should be made consistent.","section":"Sec. 3.3"},{"comment":"For the ZZ -> 4tau background, the text reports a total cross section of ~18 fb and then O(50) events after selections; this implies an acceptance of about 0.1%, which should be stated explicitly or supported by a citation.","section":"Sec. 5.1"},{"comment":"The Delta R distribution in Fig. 4 is shown only for BP2; overlaying the distribution for an unboosted a1 decay or for SM tau pairs would help demonstrate that the 82.4% threshold is a distinctive feature of the boosted topology.","section":"Fig. 4"}],"recommendation":"major_revision","confidential_remarks":"The paper is well-structured and honest about its limitations; the core simulation numbers appear internally consistent. The main problems are the overbreadth of the 'not covered by any existing search' claim and the use of unmatched literature backgrounds for 5-sigma statements. Both are fixable within the manuscript's scope, so I recommend major revision rather than rejection. The literature-coverage issue in particular is easy to address by qualifying the claim or adding a targeted survey of h->aa->4tau and light-ditau-resonance searches."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Colleague,\n\nHere's the quick take on Telba's scNMSSM di-Higgs paper. The genuinely new piece is the boosted 8tau topology for m_a1 = 8.2 GeV: the claim that 82.4% of same-a1 tau pairs have dR < 0.4 is supported by a clean truth-level distribution, and the observation that this merging is missed by standard tau tagging is a useful, concrete search recommendation. The 8b efficiency comparisons for BP1/BP3 also look consistent and the simulation chain is standard and properly described. The paper earns credit for being explicit about its own limitations: Section 5.4 lists exactly which pieces are borrowed from the literature and which are not modeled.\n\nThe weak spot, and it is a real one, is the novelty claim. The abstract says the topology is not covered by any existing ATLAS or CMS di-Higgs search; the conclusion goes further and says 'not addressed by any existing ATLAS or CMS search strategy.' Only three di-Higgs analyses are cited. No h->aa->4tau search, no low-mass ditau resonance search, no exotic-decay combination is surveyed. Some of those analyses necessarily face the same merged-tau challenge and may have partial acceptance. The simulation numbers can be right and the 'not covered' statement still false. This should be checked before the paper is used as a motivation for a new search.\n\nThe significance projections are what they say they are: indicative. Backgrounds come from published analyses with non-identical selections, no trigger or pileup model, and a flat 20% systematic that is 'illustrative.' The paper does not overclaim in the abstract — it says 'potentially compatible' — but the large Z_stat values in Table 6 will mislead a casual reader. The real message is the required background rejection, not a predicted discovery. The dependence on the companion paper for benchmarks is legitimate; those are adopted inputs from a scan, not fitted to the data at hand.\n\nWho is this for? Colleagues working on NMSSM Higgs phenomenology or on new HL-LHC trigger/object strategies. It deserves a serious referee: the central dR result is novel and reproducible, and the paper is honest. A referee should push for a broader literature check and a statement that the 'not covered' claim is limited to di-Higgs analyses with resolved tau selection. If those are fixed, this is a citable result.\n\nMy call: engage with it, but treat the 5-sigma talk as a projection and the 'no search covers this' as a hypothesis to be verified.","headline":"A genuinely new boosted 8tau channel from the scNMSSM, honestly simulated, but the 'no existing search covers this' claim outruns the citation base and the significance numbers are indicative only.","tokens_in":15846,"tokens_out":2738,"would_cite":true,"duration_ms":24748,"reading_group":"yes","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"This paper claims that the scNMSSM produces a boosted eight-tau di-Higgs signature that existing LHC searches would miss, and that an eight-bottom channel could be 5-sigma visible.","keywords":["di-Higgs production","semi-constrained NMSSM","High-Luminosity LHC","exotic Higgs decays","multi-b final states","boosted multi-tau topology","Higgs self-coupling","light pseudoscalar"],"falsifier":"A concrete test: run a full background Monte Carlo for the paper's exact object definitions at 3000 fb$^{-1}$. If the $\\ge 4b$ background exceeds about 16,000 events, or the $\\ge 3\\tau$ background exceeds about 9,100 events, the claimed 5-$\\sigma$ sensitivity for BP1/BP3 and BP2 disappears. Independently, recompute the same-$a_1$ tau-pair $\\Delta R$ distribution with an alternative shower and detector simulation; if fewer than about 82% of pairs fall below 0.4, the central merging claim of BP2 would need revision.","tokens_in":14759,"feed_emoji":"⚛️","tokens_out":10975,"duration_ms":90954,"temperature":0.7,"pith_summary":"The paper tries to show that di-Higgs production in the semi-constrained Next-to-Minimal Supersymmetric Standard Model (scNMSSM) at large $\\lambda$ and low $\\tan\\beta$ generates new final states that current LHC di-Higgs searches do not target. When the lightest pseudoscalar is heavy enough to decay to bottom pairs, the chain $h_1h_1 \\to 4a_1 \\to 8b$ produces up to eight $b$-jets, and the paper estimates signal yields around 14,000–16,000 events at 3000 fb$^{-1}$, potentially a 5-$\\sigma$ discovery if backgrounds behave as assumed. The more novel case is a pseudoscalar of 8.2 GeV that decays to taus: it is boosted by $\\gamma \\approx 7.5$, so 82.4% of same-$a_1$ tau pairs have $\\Delta R < 0.4$ and merge inside the standard jet cone. That makes the eight-tau topology invisible to resolved-tau searches and motivates a dedicated boosted-$a_1$ tagger.","feed_headline":"Eight-tau di-Higgs decays could hide inside a single jet cone","feed_subtitle":"A light pseudoscalar boosts 82% of tau pairs below the jet cone radius, opening a search channel current LHC coverage misses.","key_machinery":"The load-bearing object is the light CP-odd pseudoscalar $a_1$ of the scNMSSM, whose decay kinematics decide the final-state topology. The exotic chain $h_1 \\to a_1 a_1$ with branching ratio near 99%, followed by $a_1 \\to b\\bar{b}$ above threshold or $a_1 \\to \\tau^+\\tau^-$ below it, is what turns a standard di-Higgs event into an eight-parton or eight-tau event. For $m_{a_1} = 8.2$ GeV the important mechanism is the simple two-body boost relation $\\gamma \\approx m_{h_1}/(2m_{a_1}) \\approx 7.5$ and $\\Delta R(\\tau,\\tau) \\approx 2/\\gamma \\approx 0.27$, which predicts that the vast majority of tau pairs from one $a_1$ fit within a cone of radius 0.4. This merging is what both suppresses standard tau-tagging efficiency and creates the invariant-mass handle $m_{\\tau\\tau} \\approx m_{a_1}$ that a dedicated sub-jet tagger could exploit.","core_discovery":"The central claim, stated on the paper's own terms, is that the light-pseudoscalar regime of the scNMSSM produces a boosted eight-tau di-Higgs signature that no existing LHC di-Higgs search covers. For the benchmark with $m_{a_1} = 8.2$ GeV, the pseudoscalar receives a Lorentz boost of roughly $\\gamma = m_{h_1}/(2m_{a_1}) \\approx 7.5$, giving a characteristic tau-pair separation $\\Delta R(\\tau,\\tau) \\approx 0.27$. In the full $10^5$-event sample, 82.4% of same-$a_1$ tau pairs have $\\Delta R < 0.4$, below the standard jet cone radius, so most tau pairs are reconstructed as one merged object rather than two. The resulting $\\ge 2\\tau$ efficiency is 38.4%, about 21 times the SM $HH \\to \\tau\\tau\\tau\\tau$ baseline, but the paper finds that standard resolved-tau selection alone would not establish the signal against the estimated $Z/\\gamma^* \\to \\tau\\tau$ + jets and $t\\bar{t}$ backgrounds. For the $8b$ benchmarks, $\\varepsilon(\\ge 4b) \\approx 14\\%$ gives $S \\approx 14{,}000$ to $16{,}000$ events, which the paper describes as compatible with $5\\sigma$ sensitivity under its simplified background model with 20% systematic uncertainty.","pith_inferences":["Editorial inference: the $5\\sigma$ projections should be read as upper limits, since the background estimates are borrowed from analyses with different selections; a full background simulation could move the rates by more than 20%.","Editorial inference: the same merged-tau-pair mechanism should apply to any model with a light scalar decaying to tau pairs, not only the scNMSSM, so a boosted-$a_1$ tagger would have broad discovery power.","Editorial inference: a clean experimental test of the BP2 claim would be a search for a narrow resonance near 8.2 GeV in the invariant mass of sub-jets inside fat jets; the $ZZ \\to 4\\tau$ background has no such peak.","Editorial inference: the paper's $\\Delta R$ distribution implies that tau-tagging efficiency rises steeply with the cone radius; a scan over cone sizes between 0.2 and 0.6 would map the merging fraction and test the prediction directly."],"forward_implications":["Existing LHC di-Higgs searches, which assume resolved $b$ or $\\tau$ objects, would not count most BP2 events, so a search built on boosted-$a_1$ sub-jet tagging is needed to see this topology.","The $8b$ chain yields around 14,000–16,000 signal events at 3000 fb$^{-1}$, so a $5\\sigma$ observation would be within reach if the true background in the $\\ge 4b$ region stays within the quoted 3,000–6,000 events.","The $\\kappa_\\lambda$ suppression of 5–12% below the SM value is a shared prediction of all four benchmarks, implying that a HL-LHC measurement of the Higgs self-coupling would see a deficit in this model.","Because $h_1 \\to a_1 a_1$ dominates the Higgs width here, ordinary $HH \\to b\\bar{b}b\\bar{b}$ searches would see this model only through the small SM-like fraction, making dedicated searches for the exotic topologies essential."],"supporting_citations":[{"why":"provides the full scan of 10^6 scNMSSM points, the 66 viable SM-like points, the kappa_lambda range, and the non-resonant di-Higgs cross sections that define the four benchmarks.","marker":"[1]"},{"why":"establishes exotic h->aa decays as a source of multi-particle final states from the 125 GeV Higgs, motivating the h1h1->4a1 chains studied here.","marker":"[12]"},{"why":"supplies the four-b di-Higgs search whose background estimates and resolved-jet assumptions anchor the 8b channel and the BP4 cross-check.","marker":"[15]"},{"why":"provides the bbtautau search context used to estimate multi-tau backgrounds and to argue that the boosted 8tau topology is not covered.","marker":"[16]"},{"why":"provides the matrix-element generator used to produce gg->h1h1 signal events for all benchmark points.","marker":"[17]"},{"why":"provides the parton shower and hadronization program that handles the injected a1 decays and produces final-state particles.","marker":"[18]"},{"why":"provides the fast detector simulation used to reconstruct jets, b-tags, tau tags, and kinematic observables.","marker":"[19]"},{"why":"provides the parametrization used to normalize the gg->h1h1 signal cross section for each benchmark point.","marker":"[24]"},{"why":"supplies the bbbb search background range adopted for the 8b significance estimates and the BP4 cross-check.","marker":"[29]"},{"why":"sets the HL-LHC Higgs-pair and kappa_lambda sensitivity projections against which BP4 and the self-coupling suppression are compared.","marker":"[6]"}],"fun_headline_variants":["Eight taus in one jet: new di-Higgs channel for HL-LHC","Boosted 8-tau Higgs decays merge below jet cone radius","LHC blind spot: 8-tau di-Higgs in a single cone","New Higgs signature: 8 taus packed into one jet","Missed by ATLAS/CMS: 8-tau merged di-Higgs signal"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The sensitivity claims assume that background rates taken from published LHC analyses remain valid under the paper's different event selections, with only a 20% systematic uncertainty; if a proper background simulation puts the $\\ge 4b$ background above about 16,000 events or the $\\ge 3\\tau$ background above about 9,100 events, the 5-$\\sigma$ projections fail.","fun_headline_variants_meta":{"raw":{"variants":["Eight taus in one jet: new di-Higgs channel for HL-LHC","Boosted 8-tau Higgs decays merge below jet cone radius","LHC blind spot: 8-tau di-Higgs in a single cone","New Higgs signature: 8 taus packed into one jet","Missed by ATLAS/CMS: 8-tau merged di-Higgs signal"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000713,"raw_usage":{"total_tokens":3408,"prompt_tokens":1345,"completion_tokens":2063,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":961,"completion_tokens_details":{"reasoning_tokens":1960}},"tokens_in":961,"tokens_out":2063,"duration_ms":15034,"temperature":1.0,"reasoning_tokens":1960,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-12T11:29:15.898244+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"A concrete test: run a full background Monte Carlo for the paper's exact object definitions at 3000 fb$^{-1}$. If the $\\ge 4b$ background exceeds about 16,000 events, or the $\\ge 3\\tau$ background exceeds about 9,100 events, the claimed 5-$\\sigma$ sensitivity for BP1/BP3 and BP2 disappears. Independently, recompute the same-$a_1$ tau-pair $\\Delta R$ distribution with an alternative shower and detector simulation; if fewer than about 82% of pairs fall below 0.4, the central merging claim of BP2 would need revision.","supporting_citations":[{"cited_title":null,"cited_arxiv_id":null,"evidence_quote":"provides the full scan of 10^6 scNMSSM points, the 66 viable SM-like points, the kappa_lambda range, and the non-resonant di-Higgs cross sections that define the four benchmarks."}],"review_version":1}