{"total":19,"items":[{"citing_arxiv_id":"2606.28659","ref_index":10,"ref_count":1,"confidence":0.98,"is_internal_anchor":true,"paper_title":"Transformer-Based Active Learning for Data-Efficient Vaccine Epitope Selection in PRRS","primary_cat":"q-bio.BM","submitted_at":"2026-06-27T00:29:29+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":3.0,"formal_verification":"none","one_line_summary":"Transformer models under active learning classify high-binding epitopes from a small docking dataset more accurately than random sampling or other architectures in low-data regimes for PRRS.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2606.09889","ref_index":9,"ref_count":1,"confidence":0.98,"is_internal_anchor":true,"paper_title":"Optuna Constrained Tree-Structured Parzen Estimator Is a Joint Density Generalization of c-TPE","primary_cat":"cs.LG","submitted_at":"2026-06-03T11:16:59+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":6.0,"formal_verification":"none","one_line_summary":"Optuna's constrained TPE is joint c-TPE, the same expected constrained improvement acquisition function computed from a joint likelihood instead of an independence assumption.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2605.28478","ref_index":16,"ref_count":1,"confidence":0.98,"is_internal_anchor":true,"paper_title":"Towards Autonomous Commissioning of Industrial Drives via Multi-Objective Bayesian Optimization","primary_cat":"eess.SY","submitted_at":"2026-05-27T13:38:57+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":5.0,"formal_verification":"none","one_line_summary":"Multi-objective Bayesian optimization with TPE tunes industrial drive current controllers to expert-level performance in minutes on real hardware without a model or firmware changes.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2605.20119","ref_index":33,"ref_count":2,"confidence":0.98,"is_internal_anchor":true,"paper_title":"Toto 2.0: Time Series Forecasting Enters the Scaling Era","primary_cat":"cs.LG","submitted_at":"2026-05-19T17:08:24+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":5.0,"formal_verification":"none","one_line_summary":"Time series foundation models scale under a single training recipe, with forecast quality improving from 4M to 2.5B parameters and new SOTA results on BOOM, GIFT-Eval, and TIME benchmarks.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2605.18202","ref_index":93,"ref_count":1,"confidence":0.9,"is_internal_anchor":true,"paper_title":"Concise and Logically Consistent Conformal Sets for Neuro-Symbolic Concept-Based Models","primary_cat":"cs.LG","submitted_at":"2026-05-18T10:43:02+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":6.0,"formal_verification":"none","one_line_summary":"COCOCO is a conformal framework for NeSy-CBMs that jointly conformalizes concepts and labels, reconciles them via deduction-abduction revision, and satisfies consistency, coverage, and conciseness while retaining distribution-free guarantees.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2605.14055","ref_index":29,"ref_count":1,"confidence":0.9,"is_internal_anchor":true,"paper_title":"PEML: Parameter-efficient Multi-Task Learning with Optimized Continuous Prompts","primary_cat":"cs.CL","submitted_at":"2026-05-13T19:25:56+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":6.0,"formal_verification":"none","one_line_summary":"PEML co-optimizes continuous prompts and low-rank adaptations to deliver up to 6.67% average accuracy gains over existing multi-task PEFT methods on GLUE, SuperGLUE, and other benchmarks.","context_count":1,"top_context_role":"background","top_context_polarity":"background","context_text":"j∗ ≥1−ϵ for small ϵ >0 , then ∥α∗ −ˆα∥1 = 2(1−α ∗ j∗)≤2ϵ , and since∥ · ∥ 2 ≤ ∥ · ∥ 1: |f(θ, α ∗)−f(θ,ˆα)| ≤2Lϵ.(28) Proof. Equation (27) is the Lipschitz inequality from Assumption 1 applied to α. For (28): let j∗ = arg maxj α∗ j . The discrete solution setsˆαj∗ = 1,ˆαj = 0forj̸=j ∗. Sinceα ∗ ∈∆: ∥α∗ −ˆα∥1 = (1−α ∗ j∗) + X j̸=j ∗ α∗ j = 2(1−α ∗ j∗)≤2ϵ.(29) Applying∥ · ∥ 2 ≤ ∥ · ∥ 1 to (27) gives (28). The entropy regularization term λR(α) in (4) penalizes high-entropy distributions, driving ϵ→0 during training. Table 18 confirms this empirically: the discretization gap is 0.7 percentage points for Softmax+Argmax, versus1.1for STE, consistent with a well-peakedα ∗ and negligibleϵ. 7.3 Task Sensitivity Analysis"},{"citing_arxiv_id":"2605.11885","ref_index":65,"ref_count":1,"confidence":0.9,"is_internal_anchor":true,"paper_title":"From Clever Hans to Scientific Discovery: Interpreting EEG Foundational Transformers with LRP","primary_cat":"cs.AI","submitted_at":"2026-05-12T09:59:14+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":6.0,"formal_verification":"none","one_line_summary":"LRP on EEG transformers reveals Clever Hans artifacts in motor imagery tasks and a recurring central electrode cluster as a candidate sensorimotor signature of arousal.","context_count":1,"top_context_role":"background","top_context_polarity":"background","context_text":"[63] Raven User Guide - Technical Documentation.URL: https : / / docs . mpcdf . mpg . de / doc / computing/raven-user-guide.html#system-overviewAccessed: July 24, 2025. [64] Takuya Akiba et al. \"Optuna: A next-Generation Hyperparameter Optimization Framework\". In:Pro- ceedings of the 25th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining. 2019, pp. 2623-2631. [65] Shuhei Watanabe.Tree-Structured Parzen Estimator: Understanding Its Algorithm Components and Their Roles for Better Empirical Performance. Version 3. May 26, 2023.DOI: 10.48550/arXiv.2304.11127. arXiv:2304.11127 [cs]. Pre-published. [66] Viper-GPU User Guide - Technical Documentation.URL: https://docs.mpcdf.mpg.de/doc/ computing/viper-gpu-user-guide."},{"citing_arxiv_id":"2605.09022","ref_index":48,"ref_count":1,"confidence":0.9,"is_internal_anchor":true,"paper_title":"Inferring identified hadron production in $pp$ collisions with physics-informed machine learning at the LHC","primary_cat":"hep-ph","submitted_at":"2026-05-09T16:02:30+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":5.0,"formal_verification":"none","one_line_summary":"A physics-informed neural network infers pT spectra of pi, K, p, Lambda, and Ks in unmeasured rapidity regions from PYTHIA8 pp collisions at 13.6 TeV, achieving 1.5-5.83% yield uncertainties while reproducing yield ratios and freeze-out parameters.","context_count":1,"top_context_role":"background","top_context_polarity":"background","context_text":"Bengio, arXiv:1211.5063. [44] P. J. Siemens and J. O. Rasmussen, Phys. Rev. Lett.42, 880 (1979). [45] A. Ortiz Velasquez, P. Christiansen, E. Cuautle Flores, I. Maldonado Cervantes and G. Pai' c, Phys. Rev. Lett. 111, 042001 (2013). [46] F. Retiere and M. A. Lisa, Phys. Rev. C70, 044907 (2004). [47] J. Adamet al.(ALICE Collaboration), Nature Phys.13, 535 (2017). [48] S. Acharyaet al.(ALICE Collaboration), Phys. Rev. C 99, 024906 (2019). [49] B. B. Abelevet al.(ALICE Collaboration), Phys. Lett. B727, 371 (2013). [50] P. Braun-Munzinger, J. Stachel, J. P. Wessels and N. Xu, Phys. Lett. B344, 43 (1995). [51] P. Huovinen, P. F. Kolb, U. W. Heinz, P. V. Ruuskanen and S. A. Voloshin, Phys. Lett. B503, 58 (2001). [52] S."},{"citing_arxiv_id":"2605.06454","ref_index":62,"ref_count":1,"confidence":0.9,"is_internal_anchor":true,"paper_title":"ORTHOBO: Orthogonal Bayesian Hyperparameter Optimization","primary_cat":"cs.LG","submitted_at":"2026-05-07T15:49:03+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":5.0,"formal_verification":"none","one_line_summary":"OrthoBO introduces an orthogonal acquisition estimator subtracting an optimally weighted score-function control variate to reduce Monte Carlo variance, preserve the acquisition target, and improve ranking stability in Bayesian hyperparameter optimization.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2604.20707","ref_index":27,"ref_count":1,"confidence":0.9,"is_internal_anchor":true,"paper_title":"Generative Flow Networks for Model Adaptation in Digital Twins of Natural Systems","primary_cat":"cs.LG","submitted_at":"2026-04-22T15:48:48+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":6.0,"formal_verification":"none","one_line_summary":"GFlowNets sample multiple valid mechanistic simulator configurations for digital twin adaptation, recovering main parameter regions and preserving uncertainty in a tomato model case study.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2604.18043","ref_index":13,"ref_count":1,"confidence":0.9,"is_internal_anchor":true,"paper_title":"Optimizing Memory Allocation in Distributed Clusters with Predictive Modeling","primary_cat":"cs.DC","submitted_at":"2026-04-20T10:09:08+00:00","verdict":"CONDITIONAL","verdict_confidence":"MODERATE","novelty_score":4.0,"formal_verification":"none","one_line_summary":"A quantile-regression ensemble with safety factor reduces under-allocated jobs from 4.17% to 2.89% and average overallocation from 148% to 44.51% on SAP build data.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2604.17366","ref_index":71,"ref_count":1,"confidence":0.9,"is_internal_anchor":true,"paper_title":"ArgBench: Benchmarking LLMs on Computational Argumentation Tasks","primary_cat":"cs.CL","submitted_at":"2026-04-19T10:23:41+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":8.0,"formal_verification":"none","one_line_summary":"ArgBench unifies 33 existing datasets into a standardized benchmark for testing LLMs across 46 argumentation tasks and analyzes the impact of prompting techniques and model factors on performance.","context_count":1,"top_context_role":"method","top_context_polarity":"use_method","context_text":", 2022) in either of two training settings: Training on All Other TasksWe fine-tuned each open-weight LLM on all tasks except for the tar- get task using the respective instructions. 3 Then, we evaluated it on the target task. On the valida- tion task, we optimized the learning rate and early stopping threshold using the Tree-structured Parzen Estimator (Watanabe, 2023). Details on hyperpa- rameter optimization are in Appendix A.3, includ- ing tested ranges (Table 14) and the best values (Table 18). Training on the Target TaskTo assess the gen- eralizability of LLMs to unseen tasks, we evaluate LLMs on each target task, after fine-tuning them 3For lack of resources, we excluded Llama-3.3-70b and Mixtral-8x7b in the leave-one-task-out-experiments."},{"citing_arxiv_id":"2604.08586","ref_index":47,"ref_count":1,"confidence":0.9,"is_internal_anchor":true,"paper_title":"FluidFlow: a flow-matching generative model for fluid dynamics surrogates on unstructured meshes","primary_cat":"cs.LG","submitted_at":"2026-03-30T10:08:20+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":6.0,"formal_verification":"none","one_line_summary":"FluidFlow uses conditional flow-matching with U-Net and DiT architectures to predict pressure and friction coefficients on airfoils and 3D aircraft meshes, outperforming MLP baselines with better generalization.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2511.21300","ref_index":23,"ref_count":1,"confidence":0.9,"is_internal_anchor":true,"paper_title":"Data-Driven Reduction of Fault Location Errors in Onshore Wind Farm Collectors","primary_cat":"eess.SY","submitted_at":"2025-11-26T11:48:13+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":4.0,"formal_verification":"none","one_line_summary":"A Gated Residual Network correction model reduces fault location error by 76% in simulated onshore wind farm collector networks compared to state-of-the-art methods.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2510.02107","ref_index":9,"ref_count":1,"confidence":0.9,"is_internal_anchor":true,"paper_title":"PENEX: AdaBoost-Inspired Neural Network Regularization","primary_cat":"cs.LG","submitted_at":"2025-10-02T15:13:02+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":6.0,"formal_verification":"none","one_line_summary":"PENEX is a new formulation of the multi-class exponential loss for neural networks that supports first-order optimization and improves generalization in low-data regimes.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2509.09505","ref_index":72,"ref_count":1,"confidence":0.9,"is_internal_anchor":true,"paper_title":"Combating the Memory Walls: Optimization Pathways for Long-Context Agentic LLM Inference","primary_cat":"cs.AR","submitted_at":"2025-09-11T14:49:50+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":5.0,"formal_verification":"none","one_line_summary":"PLENA introduces a co-designed system with three optimization pathways for long-context agentic LLM inference, claiming up to 2.23x throughput over A100 and 4.04x energy efficiency.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2509.08667","ref_index":36,"ref_count":1,"confidence":0.9,"is_internal_anchor":true,"paper_title":"Minimal Data, Maximum Clarity: A Heuristic for Explaining Optimization","primary_cat":"cs.SE","submitted_at":"2025-09-10T15:03:53+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":4.0,"formal_verification":"none","one_line_summary":"EZR combines active Naive Bayes sampling and decision-tree distillation to reach over 90% of best-known multi-objective optimization performance on 60 datasets while producing clearer explanations than LIME, SHAP or BreakDown.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2503.01605","ref_index":33,"ref_count":1,"confidence":0.9,"is_internal_anchor":true,"paper_title":"A Leaf-Level Dataset for Soybean-Cotton Detection and Segmentation","primary_cat":"cs.CV","submitted_at":"2025-03-03T14:41:06+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":6.0,"formal_verification":"none","one_line_summary":"A new leaf-instance dataset for soybean-cotton detection and segmentation collected across growth stages and conditions from commercial farms is presented and validated with YOLOv11.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2409.07609","ref_index":58,"ref_count":1,"confidence":0.9,"is_internal_anchor":true,"paper_title":"Survival of the Cheapest: Cost-Aware Hardware Adaptation for Adversarial Robustness","primary_cat":"cs.CR","submitted_at":"2024-09-11T20:43:59+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":5.0,"formal_verification":"none","one_line_summary":"A decision-support framework applies AFT models to show Nvidia L4 GPUs yield 20% longer adversarial survival time at 75% lower cost than V100, with inference latency as the strongest robustness predictor.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null}],"limit":50,"offset":0}