{"id":"09648d81-e68c-4754-96e2-fd33c6d83b22","arxiv_id":"2606.05410","paper_version":1,"verdict":"UNVERDICTED","confidence":"LOW","novelty_score":4.0,"correctness_risk":"unknown","formal_verification":"none","parameter_count":0,"one_line_summary":"BSMArt version 2 adds new scanning algorithms including Affine MC, MLScanner, and CMA-ES variants to simplify and accelerate parameter space exploration in new physics models, demonstrated on soft lepton excess searches at the LHC.","lead":"BSMArt 2 updates a software tool for scanning parameter spaces in beyond-standard-model physics by adding new Monte Carlo and machine learning algorithms such as CMA-ES. A smart generalist might read it to see how computational methods are being refined to help test new physics ideas against collider data.","discovery_kind":"new_method","skeptic_critique":{"model":"grok-4.3","headline":"No quantitative benchmarks or comparisons provided for CMA-ES efficiency, diversity, or coverage vs. baselines","rationale":"The load-bearing gap identified here is identical to the reader's weakest_assumption (absence of quantitative validation). Because the full text is referenced but not reproduced in the query, the same information deficit persists; the verdict therefore remains UNVERDICTED pending the missing benchmarks.","tokens_in":1613,"tokens_out":314,"duration_ms":13174,"concrete_test":"Reproduce the two CMA-ES scans on the identical model and likelihood; report (i) valid points per CPU-hour, (ii) standard deviation of the sampled parameters, and (iii) fraction of the prior volume covered, then repeat with a standard random or Metropolis-Hastings sampler using the same budget; a factor-of-two or greater improvement in any metric would support the claim.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The central claim is that two CMA-ES variants make it 'easily possible' to locate diverse, interesting points in the soft-lepton-excess parameter space. This rests on the untested premise that the new scanners deliver measurable gains in speed or coverage. The abstract states the demonstration but supplies no acceptance rates, wall-time per valid point, parameter-space volume explored, or head-to-head numbers against random sampling, MCMC, or the prior BSMArt version. Without these data the assertion that the algorithms are 'simpler and faster' remains an unverified assertion rather than a substantiated result.","agreement_with_reader":"agree"},"referee_report":{"model":"grok-4.3","summary":"The paper presents BSMArt 2, an updated lightweight scanning tool for BSM parameter spaces that includes architectural improvements, simpler installation, expanded documentation, and new algorithms (Affine MC, Contour Finding, MLScanner, DLScanner, MLS, and CMA-ES). It showcases two CMA-ES variants applied to models relevant for soft lepton excesses at the LHC and claims that these make it easily possible to locate diverse and interesting parameter points for future testing.","tokens_in":1748,"tokens_out":278,"duration_ms":19765,"significance":"A validated tool with demonstrably faster or more efficient scanning algorithms would aid exploration of new physics models by lowering barriers to finding viable parameter points. However, the central claims of simplicity and speed rest on unshown implementation details and lack any quantitative validation, which substantially reduces the assessed significance.","major_comments":[{"comment":"Abstract: the claims that the new algorithms make scans 'simpler and faster' and that 'it is easily possible to find diverse and interesting parameter points' are presented without any acceptance rates, wall-time measurements, parameter-space coverage metrics, diversity measures, or head-to-head comparisons against random sampling, MCMC, or the prior BSMArt version; these data are required to substantiate the central performance assertions.","section":null}],"minor_comments":[],"recommendation":"major_revision","confidential_remarks":null},"author_rebuttal":{"model":"grok-4.3","summary":"We thank the referee for the constructive report. We address the major comment below.","responses":[{"response":"We agree that the abstract asserts performance improvements without the quantitative metrics requested. The manuscript describes the new algorithms and demonstrates their application to soft lepton excess models but does not contain acceptance rates, wall-time data, coverage metrics, or direct comparisons to random sampling, MCMC, or BSMArt 1. We will revise the abstract to remove unsubstantiated claims of simplicity and speed and add a dedicated subsection with benchmark results, including head-to-head comparisons where feasible.","revision_made":"yes","referee_comment":"Abstract: the claims that the new algorithms make scans 'simpler and faster' and that 'it is easily possible to find diverse and interesting parameter points' are presented without any acceptance rates, wall-time measurements, parameter-space coverage metrics, diversity measures, or head-to-head comparisons against random sampling, MCMC, or the prior BSMArt version; these data are required to substantiate the central performance assertions."}],"tokens_in":1165,"tokens_out":233,"duration_ms":23184,"standing_objections":[]},"desk_editor":{"model":"grok-4.3","letter":"The paper is mainly a software release note. It updates BSMArt with new scanning options (Affine MC, Contour Finding, MLScanner, DLScanner, MLS, and two CMA-ES variants) plus easier install and more examples. The concrete part is a short demonstration that the CMA-ES options can locate points in a soft-lepton-excess parameter space relevant for the LHC.\n\nWhat it does well is keep the package lightweight and add documentation. For users already running BSM scans, having these algorithms inside one interface can save some scripting time. The example application shows the tool can be pointed at a real LHC-motivated question without extra setup.\n\nThe soft spot is the missing evidence. The abstract states that the new methods make scans simpler and faster and that interesting points are easy to find, yet the text supplies no wall-clock times, acceptance fractions, coverage metrics, or head-to-head runs against the previous BSMArt version or plain random sampling. Without those numbers the performance claim stays untested.\n\nThis is for phenomenologists who already use or might adopt BSMArt for model scans. A reader looking for a documented, ready-to-run package with ML and evolutionary options could pick up useful details from the examples. The work is coherent on its own terms as a tool description, so it deserves a referee to check the code and request basic benchmarks.\n\nI would send it to peer review.","headline":"BSMArt 2 adds a few standard scanners to an existing tool but gives no numbers to back the 'simpler and faster' claim.","tokens_in":2247,"tokens_out":360,"would_cite":false,"duration_ms":22407,"reading_group":"maybe","serious_thinker":"yes","would_accept_peer_review":true},"rs_alignment":null,"lean_confirmation":null,"pith_extraction":{"msc":[],"pacs":[],"model":"grok-4.3","headline":"An updated parameter scanning tool adds CMA-ES and other algorithms that make it easier to find diverse testable points in new physics models.","keywords":["parameter scanning","new physics","LHC phenomenology","evolution strategy","machine learning","soft lepton excesses","model exploration"],"falsifier":"Performing side-by-side tests on a benchmark new physics model, measuring the time to find a set number of distinct viable points and the variety of those points using both the new algorithms and conventional scanning techniques.","tokens_in":2519,"feed_emoji":"","tokens_out":582,"duration_ms":26864,"temperature":0.7,"pith_summary":"This paper presents an updated version of a tool for exploring parameter spaces in theories beyond the Standard Model. It adds several new scanning algorithms based on machine learning and Monte Carlo methods, along with simpler setup and more documentation. The authors apply two versions of a covariance matrix adaptation evolution strategy scan to models that might account for soft lepton excesses at the Large Hadron Collider. A sympathetic reader would care because these improvements could allow quicker identification of model parameters worth testing in experiments. If correct, this lowers the effort required to generate interesting scenarios for future data analysis.","feed_headline":"New scanning methods locate diverse LHC-testable points in new physics models","feed_subtitle":"An updated tool incorporates evolution strategy and machine learning algorithms to speed up searches for interesting parameter spaces.","key_machinery":"The covariance matrix adaptation evolution strategy algorithm variants, used to optimize searches through the parameter space of new physics models.","core_discovery":"The updated tool incorporates new algorithms including Affine Monte Carlo, Contour Finding, machine learning scanners, deep learning scanners, and covariance matrix adaptation evolution strategy methods. Two variants of the evolution strategy scans are used to identify diverse parameter points in models relevant to soft lepton excesses at the collider, showing that such points can be found readily.","pith_inferences":["These scanning improvements might enable broader exploration of model spaces that were previously too computationally intensive.","The techniques could apply to parameter searches in other areas of particle physics beyond collider anomalies.","Combining the tool with automated model generation systems might accelerate the full cycle from model building to testing."],"forward_implications":["The new methods allow efficient location of parameter points that could explain observed anomalies in collider data.","Architectural changes make the tool easier to install and use with added examples.","Multiple scanning options give users choices suited to different exploration tasks.","Demonstrations with lepton excess models illustrate practical applications in high-energy physics."],"fun_headline_variants":["BSMArt 2 adds CMA-ES for diverse LHC parameter points","Affine MC and ML scanners update BSMArt 2 parameter searches","CMA-ES variants map soft lepton points in BSM models","New algorithms accelerate BSMArt 2 parameter space scans"],"cache_read_input_tokens":2112,"weakest_assumption_plain":"The newly added scanning algorithms deliver gains in speed, diversity of found points, or coverage compared to previous approaches.","fun_headline_variants_meta":{"raw":{"variants":["BSMArt 2 adds CMA-ES for diverse LHC parameter points","Affine MC and ML scanners update BSMArt 2 parameter searches","CMA-ES variants map soft lepton points in BSM models","New algorithms accelerate BSMArt 2 parameter space scans"]},"model":"grok-4.3","cost_usd":0.004194,"raw_usage":{"total_tokens":2047,"prompt_tokens":524,"num_sources_used":0,"completion_tokens":69,"cost_in_usd_ticks":41937000,"prompt_tokens_details":{"text_tokens":524,"audio_tokens":0,"image_tokens":0,"cached_tokens":256},"completion_tokens_details":{"audio_tokens":0,"reasoning_tokens":1454,"accepted_prediction_tokens":0,"rejected_prediction_tokens":0}},"tokens_in":524,"tokens_out":69,"duration_ms":13894,"temperature":1.0,"reasoning_tokens":1454,"cache_read_input_tokens":256,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-06-28T05:00:40.252078+00:00","model_set":{"reader":"grok-4.3"},"falsifier":"Performing side-by-side tests on a benchmark new physics model, measuring the time to find a set number of distinct viable points and the variety of those points using both the new algorithms and conventional scanning techniques.","supporting_citations":[],"review_version":1}