{"id":"3e40103c-abe2-42a6-88ac-77d6816fa4de","arxiv_id":"2412.14303","paper_version":1,"verdict":"UNVERDICTED","confidence":"HIGH","novelty_score":2.0,"correctness_risk":"unknown","formal_verification":"none","parameter_count":0,"one_line_summary":"A CMS conference proceedings reviews ttW measurement challenges and proposes a framework for a future differential cross-section measurement.","lead":"This document is a conference talk summary from the CMS collaboration about a rare process: top quark pairs produced together with a W boson. It reviews why this process is hard to measure and outlines a plan for a future, more detailed cross-section measurement.","discovery_kind":"review","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The proposal is honest, but its central feasibility claim rests on an unvalidated maximum-likelihood unfolding in §3.1; without a closure or simulation study, the framework's reliability is not established.","rationale":"The reader's verdict is UNVERDICTED, and my stress-test does not move that verdict. The paper explicitly presents only a framework proposal for a future differential ttW measurement, not a validated analysis. The single most load-bearing assumption is that the maximum-likelihood unfolding strategy of §3.1 will be unbiased and stable once implemented. That assumption is entirely untested here: there is no closure test, no simulation study, no estimate of response-matrix conditioning, and no evaluation of migration rates or profiling behavior. My concern is therefore not that the paper makes a false claim, but that its central feasibility claim is carried by an unvalidated technical step. The concrete closure test proposed above would settle whether the concern lands: if it passes, the framework is credible; if it fails, the proposed measurement would need a different unfolding scheme. Because the paper is a conference-proceedings outline and makes no testable quantitative prediction, the appropriate verdict remains UNVERDICTED, matching the reader's assessment.","tokens_in":3613,"tokens_out":3212,"duration_ms":28880,"concrete_test":"Run an MC closure test: draw pseudo-data from an independent NLO+PS ttW sample not used for the response matrix, inject a steeply falling or peaking truth distribution for a chosen variable (e.g., pT(top)), then apply the §3.1 maximum-likelihood unfolding at the 138 fb^-1 expected event yield. Check that the unfolded truth-bin pulls are compatible with the quoted uncertainties and that the covariance matrix is positive definite; also compare against a standard regularized unfolding on the same pseudo-data. If bias exceeds the statistical uncertainty or the fit is unstable, the framework as proposed cannot support a reliable differential measurement.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The paper's central claim is an outline: a future differential ttW measurement is feasible by reusing the inclusive analysis ingredients and encoding the response matrix into a maximum-likelihood fit, with migrations handled by uncertainty profiling (§3.1). This is a modest but load-bearing assertion, because the utility of the entire roadmap depends on that unfolding procedure being unbiased and stable in a low-purity phase space: the same-sign dilepton region has substantial reducible backgrounds and limited signal purity, as the author notes in §2.2 and Fig. 4. No closure test, simulation study, or validation of the ML-unfolding is presented. The sentence 'migrations are controlled by the uncertainty profiling procedure' is an assertion, not a demonstration; profiling systematic parameters can absorb real signal migrations when the MC truth model is imperfect, biasing the unfolded spectrum. In addition, the proposal to 'split the NN distribution at the reconstructed level based on the variable of interest' to diagonalise the unfolding has no quantitative evaluation of the resulting response matrix condition number or per-bin migration rates. The central claim is not internally inconsistent or contradicted by data; it is simply unsupported. Since the paper self-describes as a proposed framework (abstract) and 'ongoing' (conclusion), the honest reading is that the feasibility claim is conditional on future validation.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"This proceedings paper reviews the challenges in measuring top-quark pair production in association with a W boson (t-tbar-W) at the LHC, summarizes the recent CMS inclusive cross-section measurement, and proposes a strategy for a future differential measurement using a maximum-likelihood-based unfolding approach. The paper reports no new quantitative results; its central contribution is the proposed framework outlined in Section 3.1.","tokens_in":3852,"tokens_out":3131,"duration_ms":26534,"significance":"If the proposed framework were validated, it would provide a concrete roadmap for a differential t-tbar-W measurement and could help reconcile the observed tension between measurements and NNLO predictions. The paper honestly summarizes the known theoretical and experimental difficulties and gives proper credit to the original analyses. Its strengths are the clear presentation of the state of the art and the identification of the key ingredients needed for a differential measurement. However, the central feasibility claim is not yet supported by simulation or closure tests, so the paper's significance is conditional.","major_comments":[{"comment":"The central claim that a maximum-likelihood unfolding strategy can control response-matrix migrations via uncertainty profiling is not validated. No closure test, pseudo-experiment, or simulation study is presented, and no reference is given to a previous analysis that successfully used this exact method in a similar low-purity phase space. Profiling systematic uncertainties could absorb genuine signal migrations if the MC truth model is imperfect, biasing the unfolded spectrum. The authors should demonstrate the stability and unbiasedness of the proposed procedure, for instance with ensemble tests on simulated samples, before the framework can be judged feasible.","section":"§3.1"},{"comment":"The statement that splitting the neural-network distribution at reconstructed level based on the variable of interest 'diagonalise[s] the unfolding as much as possible' is speculative. The paper does not specify the observable(s), the binning, or the resulting migration rates; in the same-sign dilepton region the signal purity is modest (Section 2.2, Fig. 4), so large bin-to-bin migrations can be expected. A quantitative evaluation, such as the condition number of the response matrix or per-bin purities and stabilities, is needed to support this claim.","section":"§3.1"},{"comment":"The framework does not address how the data-driven background estimates and their uncertainties are propagated into the differential unfolding, although the inclusive analysis relies on them. For a differential measurement, the background shapes as functions of the observable are needed; the paper does not discuss how these shapes are obtained or how their uncertainties are included in the profile likelihood.","section":"§2.2 and §3.1"}],"minor_comments":[{"comment":"The caption contains typos: 'pannel' should be 'panel' and 'bottem' should be 'bottom'.","section":"Fig. 2 caption"},{"comment":"The sentence beginning 'in order to compute the full NNLO QCD contributions' should start with a capital 'I', and the paragraph would benefit from a period after 'not yet known'.","section":"§2.1"},{"comment":"Several DOIs contain an erroneous space after '10.1007', for example 'doi:10.1007 /JHEP07(2023)219' in Ref. [2] and 'doi:10.1007 /JHEP05(2024)131' in Ref. [1].","section":"References"},{"comment":"The abbreviation 'pdf' should be capitalized as 'PDF' when referring to parton distribution functions.","section":"§2.1"}],"recommendation":"major_revision","confidential_remarks":"This is a conference proceedings contribution rather than a full research paper. The journal should consider whether the standards for a proceedings summary or a full-length article apply. If the journal expects original research validation, the manuscript is not ready; if it accepts forward-looking proposals, it should be revised to clearly flag the unvalidated nature of the method and to address the missing quantitative support for the unfolding proposal."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"Quick take: this is a conference proceedings, not a research preprint. It reviews the current ttW situation (measured cross sections above the best theory predictions), describes the experimental and theoretical challenges, and outlines a plan for a future differential measurement using maximum-likelihood unfolding. There is no new data, no validation, and no testable claim.\n\nWhat it does well: the theoretical summary is accurate and suitably cautious—the NNLO approximations are flagged as not yet trustworthy for kinematics. The experimental challenges (low signal purity in the same-sign dilepton region, reducible backgrounds, reliance on data-driven estimates) are presented honestly and match the published CMS inclusive analysis. The improved FxFx merging scheme is a nice concrete example of recent modelling progress. As a status report for a topical conference, it is useful and readable.\n\nThe soft spot is exactly where the title points: 'towards' a differential measurement. Section 3.1 is a high-level sketch. The core idea—encode the response matrix into the maximum-likelihood fit and let the uncertainty profiling handle migrations—is plausible but unvalidated. No closure test, no simulation study, no estimate of bin-to-bin migration rates or conditioning of the response matrix. The author explicitly says the work is 'ongoing,' so the proposal is conditional. The stress-test note is right that profiling systematics can absorb real signal migrations if the MC truth model is imperfect; that is a genuine concern, but the paper does not claim to have solved it, only proposed it. For a proceedings, that is acceptable; for a research paper, it would be insufficient.\n\nThere is no circularity, no invented entities, and no quantitative claim to check. The citations are the right ones: the CMS inclusive measurement [2] and the key theory papers [5,6]. Self-citation is not an issue because the paper reports CMS results and cites the actual CMS paper.\n\nBottom line: this is a decent proceedings contribution. A reader new to ttW would get a compact, accurate picture of the field and a clear statement of what a differential measurement would require. But it is not a research preprint that deserves referee time as a novel result. A journal that publishes conference proceedings could accept it after light editing; a journal expecting original research should desk-reject. My recommendation: don't send this to peer review as a research paper. If you are judging it as a proceedings talk, judge whether the summary is accurate and the proposal is clearly stated—it is.","headline":"A clear, honest proceedings talk that summarizes the ttW puzzle and sketches an unvalidated differential-measurement plan; not a research preprint, so it should not go to peer review as one.","tokens_in":4337,"tokens_out":2099,"would_cite":false,"duration_ms":19602,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":false},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"deepseek-v4-flash","headline":"The paper proposes a fit-based unfolding strategy that would let CMS Run 2 data yield differential ttW cross sections.","keywords":["top quark","ttW production","differential cross section","maximum-likelihood unfolding","CMS","LHC","same-sign dileptons","standard model"],"falsifier":"Run a closure test on simulated $\\mathrm{t\\bar{t}W}$ pseudo-experiments: unfold known-truth distributions with the proposed maximum-likelihood procedure and check whether the fitted truth-level distributions reproduce the injected truth within uncertainties. If biases or undercoverage appear, or if the fit is numerically unstable, the proposed framework's central claim fails.","tokens_in":3405,"feed_emoji":"⚛️","tokens_out":5544,"duration_ms":50134,"temperature":0.7,"pith_summary":"The paper argues that a differential measurement of top-antitop pair production in association with a W boson ($\\mathrm{t\\bar{t}W}$) is feasible at CMS using the Run 2 dataset, and it proposes a concrete framework. The idea is to extend the existing inclusive cross-section analysis, whose measured value exceeds the latest NNLO predictions, by replacing the usual separate unfolding step with a maximum-likelihood unfolding that absorbs detector migrations into the statistical fit. The author emphasizes that the same signal region, neural-network discriminator, and data-driven background estimates can be reused. If the framework works, it would provide differential cross sections to compare against improved theory predictions and help resolve the persistent tension between measurements and calculations. This is a proposal for future work, not a completed measurement.","feed_headline":"Differential ttW cross sections proposed from CMS Run 2 data","feed_subtitle":"A fit-based unfolding reuses the inclusive analysis to map the rare top-pair-plus-W process and probe its theory gap.","key_machinery":"The maximum-likelihood unfolding strategy is the load-bearing mechanism. It replaces an explicit unfolding algorithm with a statistical fit in which the detector response matrix is built into the likelihood: signal events are binned by generator-level truth, migrations between reconstructed and true bins are profiled, and the neural-network discriminant is split at reconstruction level by the differential variable to reduce correlation. The framework also relies on the inclusive analysis's same-sign dilepton selection and neural-network classifier to control backgrounds.","core_discovery":"The central claim is that the ingredients of the CMS inclusive $\\mathrm{t\\bar{t}W}$ measurement, meaning the same-sign dilepton signal region, the neural-network discriminator, and the in-situ background and uncertainty determination, can be carried over to a differential measurement. The proposed method is a maximum-likelihood unfolding: the signal is split according to generator-level truth, so the response-matrix migrations from reconstruction to truth are encoded directly in the profile likelihood, with migrations controlled through the uncertainty-profiling procedure. To keep the unfolding as diagonal as possible, the neural-network output distribution is split at reconstructed level according to the variable of interest. The paper presents this as the outline of an ongoing analysis, not yet a measured result.","pith_inferences":["If the maximum-likelihood unfolding is to be trusted, closure tests on simulated pseudo-experiments are the natural next validation step; the paper does not report them, and without such tests the framework remains a proposal rather than a demonstrated method.","The same fit-based unfolding idea could be applied to other rare top-quark processes, where low event counts make separate unfolding steps unstable and where profiling uncertainties inside the fit is especially attractive.","A differential result in a variable sensitive to the t-W scattering diagrams could discriminate between the subleading electroweak NLO corrections and the observed inclusive excess.","Testing the framework on a toy model with an exactly known response matrix would reveal whether the uncertainty-profiling procedure introduces bias or undercoverage in the unfolded distributions."],"forward_implications":["Run 2 CMS data could yield differential $\\mathrm{t\\bar{t}W}$ cross sections in kinematic variables such as jet multiplicity or top-quark transverse momentum, with systematic uncertainties propagated consistently through the fit.","The measured differential spectra would give theory a direct target for improving NNLO approximations, whose kinematic distributions are not yet trusted.","Because backgrounds and uncertainties are determined in situ, the differential result would inherit the same data-driven background control as the inclusive measurement.","The charge-split signal regions would exploit the proton PDF charge asymmetry to further constrain backgrounds.","An improved FxFx merging scheme that excludes electroweak jets from merging would reduce the merging-scale dependence and improve the modelling of extra emissions."],"supporting_citations":[{"why":"The CMS inclusive $\\mathrm{t\\bar{t}W}$ cross-section measurement on 138 fb$^{-1}$; the proposed differential framework extends this analysis.","marker":"[2]"},{"why":"The ATLAS measurement of total and differential $\\mathrm{t\\bar{t}W}$ cross sections; provides the comparison and precedent for differential results.","marker":"[1]"},{"why":"Shows large NLO corrections from supposedly subleading electroweak contributions, motivating the theory challenge discussed in the paper.","marker":"[5]"},{"why":"Provides the state-of-the-art approximate NNLO prediction with an inclusive cross section of 745.3 fb and its scale and approximation uncertainties.","marker":"[6]"},{"why":"Introduces the original FxFx merging scheme whose merging-scale dependence motivates the improved scheme.","marker":"[7]"},{"why":"Describes the improved FxFx scheme that protects electroweak jets from merging, which the paper highlights for better $\\mathrm{t\\bar{t}W}$ modelling.","marker":"[8]"}],"fun_headline_variants":["Rare ttW gets a differential cross section plan","CMS proposes unfolding to map ttW cross sections","Unfolding aims to resolve ttW theory gap","Differential ttW measurement framework from CMS Run 2"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The proposed maximum-likelihood unfolding will absorb detector-response migrations into the fit and control them by uncertainty profiling without introducing biases or instabilities; this is assumed in Section 3.1 and no closure test or simulation validation is shown.","fun_headline_variants_meta":{"raw":{"variants":["Rare ttW gets a differential cross section plan","CMS proposes unfolding to map ttW cross sections","Unfolding aims to resolve ttW theory gap","Differential ttW measurement framework from CMS Run 2"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000527,"raw_usage":{"total_tokens":2471,"prompt_tokens":801,"completion_tokens":1670,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":417,"completion_tokens_details":{"reasoning_tokens":1607}},"tokens_in":417,"tokens_out":1670,"duration_ms":10849,"temperature":1.0,"reasoning_tokens":1607,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-11T12:19:46.842061+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Run a closure test on simulated $\\mathrm{t\\bar{t}W}$ pseudo-experiments: unfold known-truth distributions with the proposed maximum-likelihood procedure and check whether the fitted truth-level distributions reproduce the injected truth within uncertainties. If biases or undercoverage appear, or if the fit is numerically unstable, the proposed framework's central claim fails.","supporting_citations":[],"review_version":1}