{"id":"44da0402-f50f-430a-a05d-0497713bac47","arxiv_id":"2501.00857","paper_version":1,"verdict":"CONDITIONAL","confidence":"MODERATE","novelty_score":6.0,"correctness_risk":"medium","formal_verification":"none","parameter_count":4,"one_line_summary":"TrackEff estimates Belle II track-reconstruction efficiency by geometrically counting drift-chamber hits, with default thresholds tuned to a Belle II tau-tau tracking efficiency measurement.","lead":"TrackEff is a new Python tool that estimates how often a particle detector records charged-particle tracks from displaced decays, using only simple geometry and no full detector simulation. It is meant for phenomenologists who need efficiency estimates for long-lived-particle searches at Belle II without building a GEANT4 model.","discovery_kind":"new_method","skeptic_critique":{"model":"deepseek-v4-flash","headline":"Validation is in-sample and prompt-only: TrackEff's hit thresholds are tuned to the same tau-tau curve used for the quoted agreement, and no displaced-track check supports the LLP use case.","rationale":"The reader's weakest assumption is exactly the load-bearing concern: the validation is performed on the same prompt tau-tau sample used for tuning, and the intended use case is displaced tracks, for which no evidence is provided. I agree that this makes the central claim conditional rather than established. The paper is otherwise transparent: it describes the geometric model, gives equations, and explicitly acknowledges that the estimate is an upper bound on efficiency because track finding and vertexing inefficiencies are assumed removable. This is a reasonable starting point for a phenomenological tool, but the missing external validation is not a minor omission—it is the difference between a fitted curve and a predictive model. The proposed test would settle the concern by comparing TrackEff to full simulation for displaced tracks. Since the reader already issued a CONDITIONAL verdict, no change in verdict is needed; the condition is the external validation described above. Minor issues such as the TrackEff/B2TrEst naming inconsistency do not affect the central claim.","tokens_in":6523,"tokens_out":6812,"duration_ms":68684,"concrete_test":"Use Belle II's GEANT4-based simulation to generate displaced charged pions or muons from a long-lived scalar with production radii 0-50 cm and |z|<100 cm, momentum 0.1-5 GeV/c, and polar angles across the CDC acceptance. Run TrackEff on the same generator-level samples and compare per-track efficiency as a function of production point and momentum. If the absolute efficiency difference between TrackEff and full simulation exceeds ~0.05 in any kinematic bin, the paper should restrict its claims to prompt tracks or model the additional inefficiencies.","verdict_should_be":"UNCHANGED","load_bearing_attack":"Section 3 tunes the two per-cell step thresholds (n_s_inner=2, n_s=10) and uses the 20-hit requirement to match the Belle II tau-tau tracking efficiency from Ref. [27]. That same measurement is then the validation shown in Fig. 2, so the agreement is in-sample and cannot establish predictive power for the intended LLP use case. The model reduces track reconstruction to geometric hit counts and assumes all pattern-recognition, combinatorial, and vertexing inefficiencies are negligible or temporary (Sec. 2.1). For displaced tracks from an LLP decay, track finding in the Belle II CDC is not equivalent to that for prompt tracks: the track may originate inside the CDC, may enter a cell at an unusual angle, and the hit pattern may be truncated; none of these regimes is validated. The paper's own limitation note that 'additional considerations may reduce the efficiency to below the TrackEff estimate' confirms the estimate is an upper bound, but the paper does not quantify how far the true efficiency can fall for displaced tracks. Without an external check, the central claim—that phenomenologists can obtain reasonable efficiency estimates for LLP searches from generator-level MC—rests on an unverified extrapolation.","agreement_with_reader":"agree"},"referee_report":{"model":"deepseek-v4-flash","summary":"The paper presents TrackEff, a Python package that estimates the charged-track reconstruction efficiency for Belle II from generator-level kinematics by counting hits in a simplified geometric model of the CDC. The model divides the CDC into superlayers and layers, steps along the helix trajectory in 1 mm arc-length increments, and requires a minimum number of steps per cell (defaults: 2 for the innermost superlayer, 10 for others) for a cell to register a hit. A track is considered reconstructed if it accumulates at least 20 such hits. The authors validate the model by comparing its efficiency with the Belle II tau-tau tracking efficiency from Ref. [27] and use this comparison to select the default step thresholds.","tokens_in":6801,"tokens_out":6240,"duration_ms":58449,"significance":"If independently validated, TrackEff would be a useful, lightweight tool for phenomenologists who need tracking-efficiency estimates for displaced-vertex searches without running full GEANT4 simulation. The paper's strengths include a clear geometric description of the CDC model, standard helix equations in the appendix, public code availability on Zenodo, and a tunable geometry that could be adapted to other trackers. However, the central validation is in-sample: the default thresholds are tuned to the same Belle II tau-tau efficiency curve that is then shown as the agreement in Fig. 2, and no independent or displaced-track validation is provided. The paper also acknowledges in Section 2.1 that track-finding, multiplicity, hard-scatter, and vertexing inefficiencies may reduce the efficiency below the TrackEff estimate, without quantifying this reduction for the intended LLP use case. These issues are load-bearing for the central claim that phenomenologists can obtain reasonable efficiency estimates from generator-level MC.","major_comments":[{"comment":"The default step thresholds (n_s_inner=2, n_s=10) are selected by comparing TrackEff's efficiency to the Belle II tau-tau tracking efficiency from Ref. [27], and the same measurement is then shown as the validation curve in Fig. 2. The agreement is therefore an in-sample result and does not establish predictive power. No holdout sample, independent measurement, or cross-validation is presented, and no agreement metric (e.g., chi-square or coverage) is given. Please add an out-of-sample check, or explicitly reframe the figure as a tuning result and provide uncertainties on the tuned parameters.","section":"Section 3, Fig. 2"},{"comment":"The paper states that 'additional considerations may reduce the efficiency to below the TrackEff estimate' and characterizes track-finding, multiplicity, hard-scatter, and vertexing inefficiencies as temporary limitations. For the intended LLP use case—tracks originating from displaced vertices—this is not a peripheral caveat: the geometric hit-counting rule has been validated only for prompt tau-decay products, and displaced tracks can enter the CDC at unusual angles, begin inside the tracking volume, or have truncated hit patterns. The assumption that these inefficiencies are 'temporary' or negligible is unquantified and unsupported. Either provide a displaced-track validation (e.g., using a GEANT4 sample with LLP decays) or explicitly restrict the claimed range of validity and treat the result as an upper bound with a quantitative discussion of how much the true efficiency may be lower.","section":"Section 2.1, second paragraph"},{"comment":"The comparison in Fig. 2 is made visually, with the statement that 'the best match is obtained' for n_s_inner=2, n_s=10. The paper does not specify how the TrackEff error bars are computed, what the statistical uncertainty of the Belle II measurement is, or what goodness-of-fit measure was used. Consequently, the choice of defaults is not reproducible, and the claim that the efficiency dependence on thresholds is 'not strong' is not quantified. Please include the statistical procedure (e.g., a defined chi-square over the shown bins) and state the sources of the error bars.","section":"Section 3"}],"minor_comments":[{"comment":"Ref. [3] is titled 'B2TrEst', while the text and abstract refer to the package as 'TrackEff'; this naming inconsistency may confuse readers who try to access the software.","section":"Ref. [3]"},{"comment":"The figure caption would benefit from stating explicitly that the 20-hit requirement is applied to the TrackEff points; currently the connection is made only in the body text.","section":"Fig. 2 caption"},{"comment":"There are several typographical and grammatical errors: 'along thez-direction' should be 'along the z-direction', 'azimutual' should be 'azimuthal', and 'TrackEff take arc-length steps' should be 'TrackEff takes arc-length steps'.","section":"Section 2.2"},{"comment":"The sentence 'The tracking efficiency is the fraction of events in which all three pions were found' could be clarified to distinguish track-level efficiency from event-level efficiency, and it should be made explicit that the TrackEff estimate in Fig. 2 uses the same event-level definition as Ref. [27].","section":"Section 3"},{"comment":"The default arc-length step Δs=1 mm is introduced, but it is not listed among the user-tunable parameters; please state whether Δs is adjustable and whether the results depend on its value.","section":"Section 2.2"}],"recommendation":"major_revision","confidential_remarks":"The manuscript fits the journal's scope as a phenomenological tools paper, but the validation weakness is substantial and needs to be addressed before it can be accepted. The in-sample tuning and the lack of displaced-track validation directly affect the central claim. The package-name inconsistency (TrackEff vs. B2TrEst) is a minor but easily fixed issue. I would ask the authors to either provide an out-of-sample validation or clearly limit the claims and quantify the unknown uncertainty."},"author_rebuttal":null,"desk_editor":{"model":"deepseek-v4-flash","letter":"TrackEff is a genuinely useful tool for a real bottleneck: phenomenologists proposing LLP searches at Belle II currently have no quick public way to estimate displaced-track efficiencies without building a GEANT4 model. The package itself is the contribution, and it's a solid one: the helicoidal trajectory math is standard, the simplified CDC geometry is clearly described, and the code is on Zenodo with tunable parameters. I also give the authors credit for being upfront that their efficiency estimate is optimistic, in the sense that real algorithms may do worse than the hit-count ideal.\n\nThe soft spot is exactly where the stress test points: the validation is in-sample. The two step thresholds and the 20-hit requirement are tuned to the Belle II tau-tau tracking efficiency from BELLE2-NOTE-PL-2020-014, and Fig. 2 compares against that same measurement. So the agreement demonstrates that the model can fit one prompt-track curve, not that it predicts efficiencies for displaced tracks. That matters because the intended use case is LLPs decaying away from the interaction point, where tracks may start inside the CDC, cross cells at unusual angles, or have truncated hit patterns. None of that is validated. The paper's own caveat that 'additional considerations may reduce the efficiency' is an acknowledgment of this, but it doesn't quantify the gap.\n\nI don't see a fundamental error in the geometry or equations, and the limitations section is honest. The in-sample tuning is not hidden; the section is literally titled 'Validation and parameter tuning.' But calling it validation is overreach. For a tool paper, that's a fixable weakness: reframe the Fig. 2 comparison as model calibration, and add a discussion of expected systematic uncertainty for displaced vertices. An external check on a few displaced-track scenarios would be better, even if it's just toy Monte Carlo.\n\nThe name inconsistency (TrackEff in the paper, B2TrEst in the Zenodo record) is a minor but real annoyance. Should be resolved.\n\nWho's this for? Phenomenologists working on LLP proposals or early detector-design studies. It deserves a serious referee, not a desk reject. I'd encourage the journal to send it out; a good referee can push for the calibration/validation reframing without killing the paper.","headline":"TrackEff fills a real gap for LLP phenomenology, but the Fig. 2 agreement is calibration, not prediction, and the displaced-track use case remains unvalidated.","tokens_in":7261,"tokens_out":2764,"would_cite":true,"duration_ms":26327,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":["29.40.Gx"],"model":"deepseek-v4-flash","headline":"This paper claims that a simplified geometric hit-counting model, with a 20-hit requirement and two per-cell step thresholds, can reproduce Belle II's track-reconstruction efficiency and can estimate displaced-track efficiency from…","keywords":["track reconstruction efficiency","long-lived particles","displaced vertices","Belle II","central drift chamber","Monte Carlo generators","detector geometry","TrackEff"],"falsifier":"Simulate the same tau-tau events used in Ref. [27] but with track production points shifted to radii between 1 and 30 cm, compute TrackEff's predicted efficiencies, and compare them with full Belle II simulation; a large overprediction in that displaced regime would show the geometric hit-counting rule fails where it is meant to be used.","tokens_in":6322,"feed_emoji":"📐","tokens_out":5657,"duration_ms":52020,"temperature":0.7,"pith_summary":"This paper introduces TrackEff, a fast software tool that estimates whether a charged particle's track would be reconstructed in a drift-chamber detector by counting detector hits along the track's geometric path. The pitch is that phenomenologists proposing long-lived-particle searches can avoid full detector simulation and still get realistic tracking-efficiency estimates from generator-level Monte Carlo, even for tracks that originate away from the collision point. The paper validates the method by reproducing the Belle II tau-tau tracking-efficiency measurement with default thresholds of 20 hits and 2 or 10 steps per cell. If the geometric rule holds, efficiency estimates for displaced-vertex searches become cheap and transparent.","feed_headline":"Geometric hit counts predict Belle II tracking efficiency","feed_subtitle":"A new package gives displaced-track efficiency estimates from generator-level Monte Carlo, no full detector simulation needed.","key_machinery":"The central object is TrackEff, a simplified drift-chamber model that divides the Belle II central drift chamber into nine radial superlayers of rectangular $(\\Delta r, \\Delta\\phi)$ cells. The mechanism that carries the argument is helical step counting: a particle's trajectory is propagated as a helix in a uniform $B_z=1.5~\\mathrm{T}$ field, sampled in $\\Delta s=1~\\mathrm{mm}$ arc-length steps, and a cell counts as a hit only if the number of steps inside it reaches a per-superlayer threshold (defaults: $n_s^{\\rm inner}=2$ in the innermost superlayer, $n_s=10$ elsewhere). The final track decision is a user-chosen hit-count cut, with 20 hits used by default. This turns reconstruction efficiency into a purely geometric counting problem.","core_discovery":"On its own terms, the paper's discovery is that a track-reconstruction efficiency curve can be generated from a purely geometric count of drift-chamber cells, without modeling ionization, electronics, pattern recognition, or vertexing. TrackEff computes the helix of each charged particle in a uniform 1.5 T field, steps along it in 1 mm increments, and registers a hit only when a cell is traversed by enough steps; averaging over the pions of $e^+e^- \\to \\tau^+\\tau^-$ events and requiring 20 hits, the predicted efficiency follows the Belle II measurement in Fig. 2 as a function of the lepton polar angle, with the closest match at $n_s^{\\rm inner}=2$ and $n_s=10$.","pith_inferences":["Editorial: the published validation covers only prompt tau-decay tracks near the interaction point, so the paper's intended LLP use extrapolates the geometric rule to large displacements and unusual angles without direct measurement.","Editorial: the paper explicitly assumes track-finding and vertexing losses are temporary limitations, so TrackEff estimates should be treated as an achievable upper bound rather than current-run performance.","Editorial: a direct extension would be to calibrate the two per-cell step thresholds against a small full-simulation sample for any new detector geometry, turning TrackEff into a general-purpose efficiency emulator.","Editorial: for soft or highly collimated displaced tracks, pattern-recognition inefficiencies that the hit count cannot see could matter; a concrete probe is to compare TrackEff-based signal predictions with existing experimental limits."],"forward_implications":["Phenomenological long-lived-particle search proposals can quote track-reconstruction efficiencies as a function of displaced production point and momentum, computed directly from generator-level events.","Scanning detector geometries becomes fast: changing layer radii, cell counts, or lengths in the simplified model gives approximate efficiency curves without building a full detector simulation.","Track-quality cuts such as a 20-hit requirement are built in as a user choice, so proposals can test how sensitive their expected sensitivity is to the hit requirement.","The same tool can be applied to other drift-chamber or solid-state tracker concepts by adjusting the geometry parameters."],"supporting_citations":[{"why":"Supplies the measured Belle II tau-tau tracking-efficiency curve that TrackEff's parameters are tuned to reproduce.","marker":"[27]"},{"why":"Provides the Belle II CDC geometry (superlayer radial widths, layer lengths, cell counts) that the simplified model uses as defaults.","marker":"[4]"},{"why":"Documents the role of hit multiplicity in the Belle II track-finding algorithm, motivating the 20-hit criterion.","marker":"[23]"},{"why":"Describes the tau-pair track-finding efficiency measurement method used in the validation study.","marker":"[28]"},{"why":"Is the open release of the TrackEff package itself, the deliverable the paper describes.","marker":"[3]"},{"why":"Supports the paper's statement that current track-finding shortfalls are temporary, justifying efficiency estimates as achievable with future algorithms.","marker":"[24]"}],"fun_headline_variants":["TrackEff predicts tracking efficiency without GEANT4","Count drift chamber hits to estimate LLP track efficiency","Simple geometry predicts tracking efficiency for long-lived particle searches","Cell-count geometry gives Belle II tracking efficiency","Predict displaced-track efficiency from detector hit geometry"],"cache_read_input_tokens":3200,"weakest_assumption_plain":"The load-bearing premise is that a charged track is reconstructed if and only if it produces at least 20 drift-chamber hits, with the per-cell step thresholds tuned to one tau-decay measurement; this rule has only been checked near the collision point, not for displaced tracks with large offsets and unusual angles.","fun_headline_variants_meta":{"raw":{"variants":["TrackEff predicts tracking efficiency without GEANT4","Count drift chamber hits to estimate LLP track efficiency","Simple geometry predicts tracking efficiency for long-lived particle searches","Cell-count geometry gives Belle II tracking efficiency","Predict displaced-track efficiency from detector hit geometry"]},"model":"deepseek-v4-flash","effort":"low","cost_usd":0.000534,"raw_usage":{"total_tokens":2524,"prompt_tokens":854,"completion_tokens":1670,"prompt_tokens_details":{"cached_tokens":384},"prompt_cache_hit_tokens":384,"prompt_cache_miss_tokens":470,"completion_tokens_details":{"reasoning_tokens":1600}},"tokens_in":470,"tokens_out":1670,"duration_ms":14116,"temperature":1.0,"reasoning_tokens":1600,"cache_read_input_tokens":384,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-08-10T22:40:28.567540+00:00","model_set":{"reader":"deepseek-v4-flash"},"falsifier":"Simulate the same tau-tau events used in Ref. [27] but with track production points shifted to radii between 1 and 30 cm, compute TrackEff's predicted efficiencies, and compare them with full Belle II simulation; a large overprediction in that displaced regime would show the geometric hit-counting rule fails where it is meant to be used.","supporting_citations":[{"cited_title":"Report No","cited_arxiv_id":null,"evidence_quote":"Supplies the measured Belle II tau-tau tracking-efficiency curve that TrackEff's parameters are tuned to reproduce."},{"cited_title":"Track Finding Efficiency in BaBar","cited_arxiv_id":"1207.2849","evidence_quote":"Describes the tau-pair track-finding efficiency measurement method used in the validation study."}],"review_version":1}