{"id":"2a918a8d-2b9e-4ef7-a6af-d152e9f90533","arxiv_id":"2608.08609","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":5.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":0,"one_line_summary":"At DUNE, the HF-CRPA+BS cross-section model combination yields the largest simulated event rates and improves projected neutrino oscillation sensitivities by 15-30% relative to the DUNE Technical Design Report tune.","lead":"This paper uses simulations to compare how different neutrino-nucleus cross-section models change DUNE's ability to measure neutrino oscillation parameters. It finds that replacing DUNE's baseline interaction model with a combination called HF-CRPA+BS could improve sensitivity to CP violation, mass ordering, and octant by 25-30% in the simulation.","discovery_kind":"extension","skeptic_critique":{"model":"deepseek-v4-flash","headline":"The reported 25–30% 'improvement' is a conditional scenario, not a robust DUNE sensitivity gain: the cross-section-model spread is treated as a known signal rather than as the systematic uncertainty the paper's own conclusion says it is.","rationale":"The paper's quantitative headline is the claimed 25–30% improvement in DUNE sensitivities with HF-CRPA+BS. Tracing that number, it is essentially the sensitivity gain from a 27–37% increase in event rate, computed with no penalty for the fact that the cross-section model is itself uncertain. The most load-bearing step is therefore not the internal consistency of the simulations (which is fine) but the interpretation of the model spread as a known signal rather than as a systematic uncertainty. The paper's own final paragraph explicitly states that the model differences will appear as large systematic errors in real data fits, which contradicts the unqualified 'improvement' language in the abstract. This is an internal inconsistency that can be settled by a concrete computation: adding a model-scale nuisance parameter to the chi-squared and seeing how much of the apparent gain survives. I expect most of the claimed improvement is absorbed, reducing the central claim to a model-dependent hypothetical rather than a robust DUNE sensitivity projection. The reader's CONDITIONAL verdict already requires validation against data and treatment of model uncertainty as a systematic; my concern reinforces those conditions but does not move the verdict to REJECT because the paper retains value as a demonstration of model dependence and its internal comparisons are sensible. I partially agree with the reader's weakest_assumption: the reader emphasizes the external reliability of HF-CRPA/BS, while I emphasize the statistical treatment of model uncertainty; both are necessary conditions for the headline to hold, but my proposed test isolates the internal methodological flaw.","tokens_in":13385,"tokens_out":9357,"duration_ms":105005,"concrete_test":"In the Section III/IV.C GLoBES setup, add a fully correlated 25% normalization nuisance parameter (or an energy-dependent shape nuisance) for signal events and recompute the CP-violation delta-chi-squared for HF-CRPA+BS versus DUNE tune; if the relative improvement drops below about 10%, the reported gain is an artifact of neglecting the cross-section-model systematic. As a stronger check, profile over all five model configurations as alternative hypotheses in the fit.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The abstract, Section IV.C, and Tables IV–V claim HF-CRPA+BS yields 25–30% improved CP-violation, mass-ordering, and octant sensitivity and 15–20% improved atmospheric-parameter precision relative to the DUNE tune. These numbers follow directly from the 27–37% event-rate increase in Table III, because in the Section III chi-squared construction the cross-section model is assumed to be exactly known: the same model populates both true and test spectra, and the DUNE TDR pull terms include no cross-section-model error. The paper's own conclusion (Section V) states that the variation among models 'lead[s] to large systematic errors when fitting the actual data' — i.e., the spread is a systematic uncertainty. When that uncertainty is profiled or marginalized, the apparent improvement is largely absorbed, so the headline number is an artifact of treating the model as fixed rather than as a nuisance. The claim would become a real sensitivity statement only under the additional, unvalidated hypothesis that HF-CRPA+BS is the true neutrino-argon cross-section.","agreement_with_reader":"partial"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper studies how alternative neutrino-nucleus cross-section models affect DUNE's projected sensitivity to standard three-flavor oscillation parameters. For the quasi-elastic region the authors compare the DUNE TDR Valencia tune with the Llewellyn-Smith and Hartree-Fock Continuum Random Phase Approximation (HF-CRPA) models in GENIE; for the resonance region they compare Rein-Sehgal with Berger-Sehgal, using the Bodek-Ritchie ground state and Bodek-Yang DIS for all configurations. They compute total cross sections, far-detector event rates, and GLoBES-based oscillation sensitivities. They find that the HF-CRPA+BS combination gives the largest total cross section and event rates, and report 25--30% improvements in CP-violation, mass-ordering, and octant sensitivities and 15--20% improvements in the precision of sin^2(theta_23) and Delta m^2_32 relative to their implementation of the DUNE TDR tune.","tokens_in":13538,"tokens_out":4027,"duration_ms":45080,"significance":"If the reported numbers were robust, the result would be important: it would show that the choice of cross-section model changes DUNE's projected sensitivity to key oscillation parameters by roughly 15--30%, a potentially large previously underestimated effect. The simulation chain is sensible and uses standard, publicly available tools (GENIE and GLoBES), and the paper is honest in stating that the model-to-model spread implies large systematic errors when fitting data. The paper also correctly isolates the QEL and RES components and shows that the effect is not restricted to one observable. However, the headline improvement is currently presented as a DUNE sensitivity gain even though it is calculated assuming the cross-section model is exactly known and identical in the true and test spectra, with no cross-section-model nuisance parameter in the fit. The quantitative claim therefore needs to be either re-derived with a realistic treatment of cross-section uncertainties or explicitly reframed as a conditional model-scenario comparison. The paper would also be strengthened by validation of the HF-CRPA and Berger-Sehgal predictions against measured neutrino-argon data.","major_comments":[{"comment":"The reported 25-30% improvement is obtained under the assumption that the cross-section model is exactly known: the same model configuration populates both the true and test spectra in the chi^2 statistic, and the pull terms do not include any cross-section-model nuisance parameter. Under this construction the improvement is a mechanical consequence of the 27-37% event-rate increase in Table III rather than a projected sensitivity gain under realistic uncertainties. The paper's own Section V states that the model spread 'lead[s] to large systematic errors when fitting the actual data,' which directly undermines the abstract's presentation of the improvement as a DUNE sensitivity result. The authors should either marginalize or profile over a cross-section-model systematic (or a conservative normalization/shape uncertainty) and report the residual improvement, or explicitly re-label the numbers as conditional idealized comparisons.","section":"Section III, Eq. (4); Tables III-IV"},{"comment":"The claim that HF-CRPA+BS is the 'best' configuration is not supported by any comparison with measured neutrino-argon cross-section data. The manuscript itself notes that HF-CRPA lacks full relativistic dynamics and is interpolated with SuSAv2 at large momentum transfer, so it is not established that this model is more accurate than the DUNE tune. Without validation against data (e.g., MicroBooNE, MINERvA, or T2K cross-section measurements), the headline improvement is conditional on an unvalidated model. I request at least a quantitative comparison of GENIE predictions with published argon cross-section data, or a clear caveat that the improvement is a model-scenario statement rather than an expected gain.","section":"Section II.A.3 and Section IV.A"},{"comment":"The comparison with the DUNE TDR baseline is not apples-to-apples: the text states that the DUNE tune's RS and BY implementations differ from the authors' RS and BY models because of different GENIE versions and 'some specific tuning used in DUNE simulations.' The paper does not specify the GENIE version used, nor how the DUNE tune was reproduced. This ambiguity affects the quantitative baseline: the 15-30% numbers are relative to an approximate DUNE tune, not necessarily the official TDR configuration. Please provide the exact GENIE version and the tuned parameter values, and justify that the residual model differences do not bias the comparison.","section":"Section IV.A, first paragraph"}],"minor_comments":[{"comment":"The table caption says 'at the parameter values' with no following period, and the row labels for beam modes and channels are inconsistent with the text (e.g., 'Neutrino (FHC)' versus 'neutrino mode').","section":"Section IV.B, Table III"},{"comment":"The axis labels and y-axis titles are garbled; for example, the CP-violation panels show '= 2 3 5' and '= 2' instead of clear chi^2 definitions. Please clean up the figures so the quantitative axes are legible.","section":"Figures 4 and 5"},{"comment":"The sentence beginning 'By combining the self-consistent mean field of HF...' is an incomplete sentence and should be finished or merged with the preceding paragraph.","section":"Section II.A.3, final sentence"},{"comment":"The paper quotes '15-20%' improvement for sin^2(theta_23) and Delta m^2_32, but Table V shows specific reductions of 19.9% and 14.9%; please state the specific values to avoid a misleading range.","section":"Abstract and Section IV.C"},{"comment":"Reference [20] is a presentation slide set; if the DUNE baseline model is cited, please also cite a peer-reviewed description (e.g., the DUNE TDR or a published tuning document) so readers can reproduce the comparison.","section":"Reference [20]"}],"recommendation":"major_revision","confidential_remarks":"The paper has a useful conditional result, but the abstract overstates robustness because the cross-section model is treated as known rather than as a systematic uncertainty. The authors' own conclusions acknowledge this. I recommend major revision rather than rejection because the issue can be addressed by re-deriving the sensitivity with a cross-section nuisance parameter or by clearly reframing the numbers as model-scenario comparisons. The lack of data validation is a related but separate concern that should be addressed, at least with a caveat, before publication."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"You should know one thing up front: this is a clean, internally consistent GENIE+GLoBES comparison of QEL/RES model combinations for DUNE, and the specific HF-CRPA+BS results are new. But the 25–30% 'improvement' is conditional on HF-CRPA+BS being the true cross-section. The stress-test concern holds up: in the chi-square construction, the same model populates both true and test spectra, and the pull terms include no cross-section-model error. So the headline numbers are scenarios, not robust sensitivity gains.\n\nWhat the paper does well: it isolates QEL and RES while fixing all other channels, which lets the comparison stay clean. The paper is honest about the GENIE version mismatch with the DUNE TDR, explicitly says the views are not those of the DUNE Collaboration, and the individual QEL and RES sensitivity plots are a useful diagnostic. The cross-section ranking they find—DUNE tune weakest in total because RS is weaker than BS—is a useful sanity check, even if model-dependent. The 27–37% event-rate increase and the corresponding sensitivity shifts are quantified in a way that is easy to follow.\n\nSoft spots, in proportion: the main one is the fixed-model assumption. Because the model is treated as exactly known, the quantitative claims are conditional. The paper's own conclusion says the model spread 'lead[s] to large systematic errors when fitting the actual data'—that admission undercuts the abstract's 'improvement' language. They also do not validate HF-CRPA or BS against measured argon cross-section data, and they provide no config files or explicit GENIE version, so exact reproduction is not possible. These are moderate issues; the comparative direction is sound.\n\nWho this is for: anyone working on DUNE oscillation sensitivities and cross-section systematics. It deserves a serious referee, but with a request to reframe the claims as conditional scenarios and ideally to add a simple profiling of the model spread as a nuisance parameter, which would turn the spread into a systematic error and change the headline.","headline":"A competent but conditional DUNE simulation study: the HF-CRPA+BS numbers are new and internally consistent, but the headline 'improvement' only holds if that model is the true cross-section.","tokens_in":14123,"tokens_out":2057,"would_cite":false,"duration_ms":24985,"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":"This paper argues that the choice of neutrino–nucleus cross-section model can change DUNE's projected oscillation sensitivities by 15–30%, with the HF-CRPA+BS combination outperforming the baseline tune.","keywords":["neutrino oscillations","DUNE long-baseline experiment","neutrino-nucleus cross-section models","quasi-elastic scattering","baryon resonance production","CP violation sensitivity","mass ordering sensitivity","GENIE event generator"],"falsifier":"Measure the total charged-current neutrino–argon cross-section in the 1–5 GeV range (or the single-pion production cross-section that separates Rein–Sehgal from Berger–Sehgal) and compare the central value with the HF-CRPA+BS prediction; if the data match the DUNE-tune prediction instead, the quoted 25–30% sensitivity improvements will not materialize.","tokens_in":13131,"feed_emoji":"⚛️","tokens_out":15341,"duration_ms":131709,"temperature":0.7,"pith_summary":"This paper argues that the choice of neutrino–nucleus cross-section model is a large, largely overlooked systematic in DUNE's oscillation program. Using the GENIE event generator and the official GLoBES simulation files, the authors compare four combinations of quasi-elastic (QEL) and resonance (RES) models against the baseline \"DUNE tune\" (Valencia QEL + Rein–Sehgal RES). They find that combining the Hartree–Fock Continuum Random Phase Approximation (HF-CRPA) for QEL with the Berger–Sehgal (BS) model for RES gives the strongest total cross-section in the few-GeV region where the DUNE flux peaks, and the DUNE tune is the weakest. In their simulated sensitivities, this model improves CP-violation, mass-ordering, and octant sensitivities by 25–30%, and the precision of $\\sin^2\\theta_{23}$ and $\\Delta m^2_{32}$ by 15–20%, relative to the DUNE tune. A sympathetic reader would take away that cross-section modeling, especially in the resonance region, can materially change DUNE's projected physics reach.","feed_headline":"Cross-section model choice shifts DUNE sensitivity by 25–30%","feed_subtitle":"The baseline DUNE tune predicts the weakest cross section; the choice of nuclear model changes DUNE's physics reach.","key_machinery":"The machinery that carries the argument is the combination of GENIE cross-section models and the GLoBES DUNE simulation. The central objects are the two alternative QEL models—Llewellyn-Smith (LS) and the Hartree–Fock Continuum Random Phase Approximation (HF-CRPA), a microscopic nuclear-response model that includes long-range correlations and continuum final states, interpolated with SuSAv2 at large momentum transfer—and the two RES models—Rein–Sehgal (RS) and Berger–Sehgal (BS), the latter adding updated resonance parameters, non-resonant background, and explicit lepton-mass corrections to RS. The paper fixes all other channels (DIS, 2p2h, ground state) to the DUNE TDR configuration and isolates QEL and RES variations, then uses the Poisson log-likelihood with pull systematics to translate event-rate differences into $\\Delta\\chi^2$ sensitivities. The mechanism is straightforward: because the DUNE flux peaks near 2.5 GeV where RES dominates, the higher total cross-section of HF-CRPA+BS increases far-detector signal events by 27–37% (Table III), and that event surplus directly inflates the test statistic for every oscillation observable.","core_discovery":"The central claim is that within DUNE's official simulation framework, the configuration using HF-CRPA for quasi-elastic scattering and the Berger–Sehgal model for resonance production yields substantially better oscillation sensitivities than the baseline DUNE tune, because it predicts a larger neutrino–argon cross-section in the energy range that dominates the DUNE flux. Concretely, the authors report that HF-CRPA+BS raises the CP-violation sensitivity from $\\Delta\\chi^2 = 78.01$ to $100.96$ at $\\delta_{\\mathrm{CP}} \\simeq 110^\\circ$ (about 29.4%), the mass-ordering sensitivity from $\\Delta\\chi^2 = 558.3$ to $705.2$ near $\\delta_{\\mathrm{CP}} \\simeq 270^\\circ$ (about 26.3%), the octant sensitivity by about 25%, and shrinks the $3\\sigma$ allowed widths of $\\sin^2\\theta_{23}$ and $\\Delta m^2_{32}$ by about 20% and 15%, respectively. The gain is driven by the resonance region: although the Valencia QEL model has the strongest quasi-elastic cross-section, the DUNE tune's Rein–Sehgal resonance model is the weakest, and RES dominates the total cross-section where the beam peaks around 2.5 GeV. The paper therefore concludes that cross-section modeling in the quasi-elastic–resonance transition region is a source of 15–30% uncertainty in DUNE's projected oscillation sensitivity, and that validating these nuclear models is essential.","pith_inferences":["If the HF-CRPA+BS cross-section were reweighted to match published neutrino–argon scattering data, the 25–30% gain might shrink or vanish, since the improvement is driven entirely by a larger predicted cross-section rather than by reduced uncertainty or better data agreement.","The authors compare against the DUNE TDR baseline but note that the current DUNE analysis uses a different tune; re-running the analysis against that current tune would show whether the model ranking persists.","A testable prediction of the paper is that model differences should appear most strongly in the low-energy part of the reconstructed spectrum, where BS's lepton-mass corrections matter; DUNE's own data, once collected, could distinguish the models by spectral shape rather than total rate.","Because the paper fixes the 2p2h and DIS treatments, the reported improvements are conditional on those choices; varying the 2p2h strength could substantially change the relative ranking of the QEL models."],"forward_implications":["If HF-CRPA+BS is the correct model, DUNE's event rates will be roughly 27–37% higher than the TDR baseline, making the experiment's planned sensitivity to CP violation and mass ordering conservative.","Since the DUNE tune is the weakest of the four configurations studied, the choice of cross-section model is not a minor technical detail: the spread across models defines an unquantified 15–30% systematic band on DUNE's headline oscillation results.","The improvement is concentrated in the resonance region, so experimental validation of single-pion production models (BS versus RS) is a direct lever on DUNE's physics reach.","Because all oscillation observables improve together, the cross-section model acts as a global enhancement of the oscillation information content rather than a channel-specific distortion."],"supporting_citations":[{"why":"Defines the Berger–Sehgal resonance model whose larger cross-section drives the sensitivity gain.","marker":"[18]"},{"why":"Defines the Rein–Sehgal resonance model used in the baseline DUNE tune that the paper finds weakest.","marker":"[17]"},{"why":"Provide the HF-CRPA quasi-elastic model that yields the best overall configuration when paired with BS.","marker":"[15, 16]"},{"why":"Supplies the Valencia quasi-elastic model used in the DUNE tune, the baseline for comparison.","marker":"[21]"},{"why":"Defines the DUNE experimental configuration and physics sensitivities that the paper's projections build on.","marker":"[12]"},{"why":"Provides the official GLoBES configuration files approximating the DUNE TDR used in the simulations.","marker":"[33]"},{"why":"Documents the GENIE implementation of HF-CRPA, including the SuSAv2 interpolation at high momentum transfer.","marker":"[29]"},{"why":"Supplies the SuSAv2 model used to extend HF-CRPA beyond its relativistic validity limit.","marker":"[28]"}],"fun_headline_variants":["Alternative nuclear model boosts DUNE sensitivity by up to 30%","Resonance model choice can raise DUNE reach by 25-30%","DUNE tune weakest: alternative cross-sections improve CP sensitivity","Better resonance model lifts DUNE's oscillation sensitivity up to 30%","Swap cross-section model to boost DUNE's CP and mass ordering sensitivity"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that GENIE's implementation of HF-CRPA and the Berger–Sehgal model reliably describe neutrino–argon interactions in DUNE's few-GeV energy range, so the larger cross-section they predict is real physics rather than a modeling artifact.","fun_headline_variants_meta":{"raw":{"variants":["Alternative nuclear model boosts DUNE sensitivity by up to 30%","Resonance model choice can raise DUNE reach by 25-30%","DUNE tune weakest: alternative cross-sections improve CP sensitivity","Better resonance model lifts DUNE's oscillation sensitivity up to 30%","Swap cross-section model to boost DUNE's CP and mass ordering sensitivity"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000993,"raw_usage":{"total_tokens":4293,"prompt_tokens":1117,"completion_tokens":3176,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":733,"completion_tokens_details":{"reasoning_tokens":3081}},"tokens_in":733,"tokens_out":3176,"duration_ms":24741,"temperature":1.0,"reasoning_tokens":3081,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-14T04:30:25.492438+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Measure the total charged-current neutrino–argon cross-section in the 1–5 GeV range (or the single-pion production cross-section that separates Rein–Sehgal from Berger–Sehgal) and compare the central value with the HF-CRPA+BS prediction; if the data match the DUNE-tune prediction instead, the quoted 25–30% sensitivity improvements will not materialize.","supporting_citations":[{"cited_title":"CP-violation or Nuclear Excitation: Reviewing the Role of Neutrino Interaction Model Uncertainties on Accelerator-Based Neutrino Oscillation Measurements","cited_arxiv_id":"2605.28671","evidence_quote":"Defines the Berger–Sehgal resonance model whose larger cross-section drives the sensitivity gain."},{"cited_title":"Long-baseline neutrino oscillation physics po- tentialoftheDUNEexperiment,","cited_arxiv_id":null,"evidence_quote":"Defines the Rein–Sehgal resonance model used in the baseline DUNE tune that the paper finds weakest."},{"cited_title":"Low-energy excitations and quasielas- tic contribution to electron-nucleus and neutrino-nucleus scattering in the continuum random-phase approxima- tion,","cited_arxiv_id":null,"evidence_quote":"Supplies the Valencia quasi-elastic model used in the DUNE tune, the baseline for comparison."},{"cited_title":"Extensions of super- scaling from relativistic mean field theory: The susav2 model,","cited_arxiv_id":null,"evidence_quote":"Provides the official GLoBES configuration files approximating the DUNE TDR used in the simulations."},{"cited_title":"Scaling and superscaling in inclusive electron–nucleus scattering,","cited_arxiv_id":null,"evidence_quote":"Documents the GENIE implementation of HF-CRPA, including the SuSAv2 interpolation at high momentum transfer."},{"cited_title":"Superscaling in inclusive electron–nucleus scattering,","cited_arxiv_id":null,"evidence_quote":"Supplies the SuSAv2 model used to extend HF-CRPA beyond its relativistic validity limit."}],"review_version":1}