{"id":"1179f11f-0d11-46e4-9cbd-c01a2cd5e66d","arxiv_id":"2607.00717","paper_version":1,"verdict":"UNVERDICTED","confidence":"LOW","novelty_score":6.0,"correctness_risk":"unknown","formal_verification":"none","parameter_count":1,"one_line_summary":"New ACE potentials achieve near-DFT accuracy on screw dislocation Peierls barriers and glide planes for Mo and W, with substantial gains for V, Nb, and Ta, while highlighting that single-humped barriers alone do not guarantee correct glide behavior.","lead":"The paper develops atomic cluster expansion potentials fitted to an extended DFT database for modeling screw dislocations in BCC refractory metals V, Nb, Ta, Mo, and W. These models reach near-DFT accuracy on Peierls barriers and glide for Mo and W and show clear gains for the others, enabling kink-pair calculations.","discovery_kind":"new_method","skeptic_critique":{"model":"grok-4.3","headline":"No significant objection identified","rationale":"Reader correctly identified the key assumption from the abstract. Full-text reference does not introduce new load-bearing gaps beyond that assumption; the paper's own statements on sensitivity to database composition align with the concern. No adjustment to UNVERDICTED is warranted without the actual manuscript content.","tokens_in":1736,"tokens_out":233,"duration_ms":15563,"concrete_test":"Recompute the Peierls barriers and glide planes for the five elements using the final ACE potentials but with the added dislocation configurations removed from the training set; if the reported accuracy drops to the level of prior potentials, the database-extension step is the controlling factor.","verdict_should_be":"UNCHANGED","load_bearing_attack":"The abstract already flags that transferability depends sensitively on database composition and that a single-humped Peierls barrier is insufficient validation. The central claim (near-DFT accuracy for Mo/W, substantial gains for V/Nb/Ta) is presented with that caveat; no internal inconsistency or unstated assumption that would invalidate the reported improvements is apparent from the provided text.","agreement_with_reader":"agree"},"referee_report":{"model":"grok-4.3","summary":"The paper claims to have developed an array of atomic cluster expansion (ACE) potentials for screw dislocations in BCC refractory metals (V, Nb, Ta, Mo, W) by extending an existing DFT database. These potentials significantly improve the description of screw dislocation properties, achieving near-DFT accuracy for Mo and W and substantial improvement for V, Nb, and Ta. The results indicate that transferability depends sensitively on database composition and element-specific energetics, and that a single-humped Peierls barrier is not sufficient validation for accurate dislocation glide prediction. For Nb, Mo, and W, the potentials enable reliable calculation of kink-pair activation enthalpies described by Kocks' law and a line-tension model.","tokens_in":1797,"tokens_out":494,"duration_ms":38968,"significance":"If the results hold, this work provides valuable interatomic potentials for modeling the complex plasticity behavior in refractory metals, which are important for high-temperature structural applications. The emphasis on the role of database composition in achieving transferability offers practical insights for developing machine learning potentials. The calculation of kink-pair activation enthalpies represents a step toward bridging atomistic simulations with macroscopic mechanical properties. The caveat about validation metrics is a strength in the presentation.","major_comments":[{"comment":"§4.2: the claim of near-DFT accuracy for Mo and W (and substantial improvement for V, Nb, Ta) is presented without quantitative error bars on the Peierls barrier or glide-plane predictions, nor an explicit validation protocol. This is load-bearing for assessing the magnitude and robustness of the central improvements.","section":"§4.2"},{"comment":"§5.1: it is not stated whether the reported glide-plane predictions were obtained before or after inspecting the target DFT data. This directly affects the independence of the transferability claims from fitting artifacts.","section":"§5.1"}],"minor_comments":[{"comment":"Figure 3: axis labels and legends could be clarified to distinguish the different ACE models more clearly.","section":"Figure 3"},{"comment":"Ensure all element-specific results are cross-referenced consistently between text, tables, and supplementary material.","section":"Tables 1-3"}],"recommendation":"major_revision","confidential_remarks":null},"author_rebuttal":{"model":"grok-4.3","summary":"We thank the referee for the constructive comments. We address each major point below and will revise the manuscript accordingly to improve clarity and robustness.","responses":[{"response":"We agree that quantitative error bars and an explicit validation protocol are needed to support the accuracy claims. In the revised manuscript we will report error bars on all Peierls barrier and glide-plane values (derived from DFT convergence tests and fitting residuals) and add a dedicated paragraph in §4.2 describing the validation protocol, including the exact comparison metrics and cross-validation steps used.","revision_made":"yes","referee_comment":"[§4.2] §4.2: the claim of near-DFT accuracy for Mo and W (and substantial improvement for V, Nb, Ta) is presented without quantitative error bars on the Peierls barrier or glide-plane predictions, nor an explicit validation protocol. This is load-bearing for assessing the magnitude and robustness of the central improvements."},{"response":"The glide-plane predictions were performed after fitting but without prior inspection of the specific target DFT glide-plane configurations used for validation. The fitting database contained only bulk, vacancy, and surface data; no glide-plane information was included, so the reported results constitute an independent transferability test. We will add an explicit statement of this sequence to §5.1 in the revision.","revision_made":"yes","referee_comment":"[§5.1] §5.1: it is not stated whether the reported glide-plane predictions were obtained before or after inspecting the target DFT data. This directly affects the independence of the transferability claims from fitting artifacts."}],"tokens_in":1438,"tokens_out":325,"duration_ms":35749,"standing_objections":[]},"desk_editor":{"model":"grok-4.3","letter":"This paper develops ACE potentials for screw dislocations in V, Nb, Ta, Mo, and W by extending an existing DFT database. The models reach near-DFT accuracy on Peierls barriers and glide for Mo and W, with clear gains for the group-5 elements, and they support kink-pair enthalpy calculations for Nb, Mo, and W that align with Kocks' law and line-tension models.\n\nWhat stands out is the explicit demonstration that a single-humped Peierls barrier is not enough to guarantee correct glide-plane behavior, and that transferability depends sensitively on database composition and element-specific energetics. They present these limits directly rather than burying them.\n\nThe improvements over classical and earlier ML potentials on dislocation-specific targets are the concrete advance. Earlier work often stopped at bulk or simpler defect properties, so targeting core energetics and glide here fills a gap.\n\nThe soft spots are the absence of quantitative error bars or a spelled-out validation protocol in the abstract, plus no statement on whether glide predictions were generated before or after seeing the target DFT data. These details matter for judging how much is robust versus fit-specific. The central claims still rest on external DFT comparisons, so circularity does not appear to be an issue.\n\nThis is for computational materials researchers who model high-temperature plasticity and need better interatomic potentials for dislocation dynamics. Anyone running kink-pair or flow-stress calculations in these metals would get direct value from the new models and the validation lessons.\n\nIt deserves peer review because it delivers usable new potentials with targeted testing on quantities that matter for the subfield.","headline":"New ACE potentials get closer to DFT on screw dislocations in refractory BCC metals, with honest caveats on database dependence.","tokens_in":2285,"tokens_out":391,"would_cite":false,"duration_ms":27469,"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":"Extending an existing DFT database produces atomic cluster expansion potentials that reach near-DFT accuracy for screw dislocation properties in BCC refractory metals.","keywords":["atomic cluster expansion","screw dislocations","BCC refractory metals","Peierls barrier","interatomic potentials","DFT database","kink-pair activation","plasticity"],"falsifier":"An independent DFT calculation, performed on a screw-dislocation configuration never included in the fitting database, that yields a Peierls barrier height or preferred glide plane differing from the ACE prediction.","tokens_in":2621,"feed_emoji":"","tokens_out":814,"duration_ms":31063,"temperature":0.7,"pith_summary":"The paper constructs a family of atomic cluster expansion potentials for the group-5 and group-6 refractory metals by adding targeted configurations to an existing density-functional-theory database. These potentials are shown to reproduce the compact three-fold cores, Peierls barriers, and glide planes of screw dislocations far more faithfully than earlier classical or machine-learned models. Near-DFT fidelity is obtained for molybdenum and tungsten; substantial gains appear for vanadium, niobium and tantalum. The authors demonstrate that reproducing a single-humped Peierls barrier is not enough to guarantee correct glide behavior, and that transferability is controlled by database composition together with element-specific energetics. For niobium, molybdenum and tungsten the models further allow direct computation of kink-pair activation enthalpies.","feed_headline":"ACE potentials reach near-DFT accuracy on BCC screw dislocations","feed_subtitle":"Models for V, Nb, Ta, Mo and W improve core structure and Peierls barriers over prior potentials; kink-pair enthalpies become accessible for","key_machinery":"Atomic cluster expansion (ACE) potentials fitted to an extended DFT database, which encode the many-body interactions that set the three-fold symmetric core structure and the height and shape of the Peierls barrier.","core_discovery":"The developed ACE potentials significantly improve the description of screw dislocation properties, achieving near-DFT accuracy for Mo and W and substantial improvement for V, Nb, and Ta. The results show that transferability to screw dislocation behavior depends sensitively on both database composition and element-specific energetics, and that achieving a single-humped Peierls barrier alone is not a sufficient validation metric for accurate prediction of dislocation glide. For Nb, Mo, and W, the developed ACE models also enable reliable calculation of kink-pair activation enthalpies, which are well described by both Kocks' law and a line-tension model.","pith_inferences":["The same fitting strategy could be tested on larger supercells that include dislocation–dislocation interactions or grain boundaries to check whether the improvement persists at engineering length scales.","Direct comparison of the predicted temperature dependence of yield stress against single-crystal experiments on Nb, Mo or W would provide an external test of the kink-pair enthalpies.","If the element-specific energetics requirement holds, the method may need re-optimization for alloys rather than pure metals."],"forward_implications":["Accurate, temperature-dependent flow-stress predictions become feasible for these refractory metals once kink-pair enthalpies are reliably obtained.","Database composition must be tuned specifically for dislocation properties; generic fitting does not guarantee transferability.","Reproducing a single-humped Peierls barrier is insufficient validation; the full glide-plane energetics must also be checked against DFT.","Element-specific differences in core energetics remain decisive even after the ACE functional form is adopted.","The same database-extension route can be applied to other BCC metals or to related defects whose core structures dominate plasticity."],"fun_headline_variants":["ACE models achieve near-DFT for BCC screw dislocations in Mo and W","Substantial improvement in screw dislocation properties via ACE potentials","ACE enables reliable kink-pair activation enthalpies for Nb Mo W","ACE transferability to screw dislocation glide sensitive to database"],"cache_read_input_tokens":2112,"weakest_assumption_plain":"Extending an existing DFT database with additional configurations is sufficient to produce transferable potentials whose dislocation behavior is controlled by the database composition and element-specific energetics rather than by fitting artifacts.","fun_headline_variants_meta":{"raw":{"variants":["ACE models achieve near-DFT for BCC screw dislocations in Mo and W","Substantial improvement in screw dislocation properties via ACE potentials","ACE enables reliable kink-pair activation enthalpies for Nb Mo W","ACE transferability to screw dislocation glide sensitive to database"]},"model":"grok-4.3","cost_usd":0.01197,"raw_usage":{"total_tokens":5183,"prompt_tokens":736,"num_sources_used":0,"completion_tokens":67,"cost_in_usd_ticks":119703000,"prompt_tokens_details":{"text_tokens":736,"audio_tokens":0,"image_tokens":0,"cached_tokens":64},"completion_tokens_details":{"audio_tokens":0,"reasoning_tokens":4380,"accepted_prediction_tokens":0,"rejected_prediction_tokens":0}},"tokens_in":736,"tokens_out":67,"duration_ms":18538,"temperature":1.0,"reasoning_tokens":4380,"cache_read_input_tokens":64,"cache_creation_input_tokens":0},"cache_creation_input_tokens":0},"created_at":"2026-07-02T10:17:28.837899+00:00","model_set":{"reader":"grok-4.3"},"falsifier":"An independent DFT calculation, performed on a screw-dislocation configuration never included in the fitting database, that yields a Peierls barrier height or preferred glide plane differing from the ACE prediction.","supporting_citations":[],"review_version":1}