{"total":14,"items":[{"citing_arxiv_id":"2606.15830","ref_index":9,"ref_count":1,"confidence":0.88,"is_internal_anchor":false,"paper_title":"MSC-CMA-ES: Structure-Aware Restarts for CMA-ES via Cyclic Nearest-Better Basin Discovery","primary_cat":"cs.NE","submitted_at":"2026-06-14T14:07:56+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":6.0,"formal_verification":"none","one_line_summary":"MSC-CMA-ES makes CMA-ES restarts structure-aware via cyclic nearest-better basin discovery on Sobol pre-samples, achieving 2.7x higher target coverage than BIPOP-CMA-ES on composition functions across CEC suites.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2606.10698","ref_index":67,"ref_count":1,"confidence":0.88,"is_internal_anchor":false,"paper_title":"Efficient AI-Inspired Reduction of Feynman Integrals via Tube Seeding","primary_cat":"hep-ph","submitted_at":"2026-06-09T10:59:00+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":8.0,"formal_verification":"none","one_line_summary":"Machine learning discovers a tube-seeding strategy for IBP reduction of Feynman integrals that scales linearly with numerator power, demonstrated on rank-20 2-loop 5-point integrals.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2606.09220","ref_index":8,"ref_count":1,"confidence":0.88,"is_internal_anchor":false,"paper_title":"Quantitative Performance Analysis of Stopping Criteria for CMA-ES","primary_cat":"cs.NE","submitted_at":"2026-06-08T08:56:38+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":4.0,"formal_verification":"none","one_line_summary":"Empirical benchmarking shows tolfunhist and the full portfolio stop CMA-ES closest to the optimal evaluation count on BBOB, while tolfun and tolfunhist often trigger before full stagnation.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2605.29560","ref_index":11,"ref_count":1,"confidence":0.88,"is_internal_anchor":false,"paper_title":"Battery-Sim-Agent: Leveraging LLM-Agent for Inverse Battery Parameter Estimation","primary_cat":"cs.AI","submitted_at":"2026-05-28T08:12:47+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":7.0,"formal_verification":"none","one_line_summary":"Battery-Sim-Agent reframes inverse battery parameter estimation as an LLM reasoning task in closed loop with a simulator and outperforms Bayesian optimization baselines on diverse benchmarks.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2605.28202","ref_index":23,"ref_count":1,"confidence":0.88,"is_internal_anchor":false,"paper_title":"Natural Functional Gradients for Smooth Trajectory Optimization","primary_cat":"cs.RO","submitted_at":"2026-05-27T09:25:57+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":6.0,"formal_verification":"none","one_line_summary":"A trajectory optimization method performs geometry-aware updates in function space via natural functional gradients and Monte-Carlo estimation on a smoothed surrogate objective to improve feasibility and smoothness in robotic manipulation.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2605.16668","ref_index":40,"ref_count":1,"confidence":0.88,"is_internal_anchor":false,"paper_title":"GraViti: Graph-Level Variational Autoencoders with Relaxed Permutation Invariance","primary_cat":"cs.LG","submitted_at":"2026-05-15T22:08:45+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":6.0,"formal_verification":"none","one_line_summary":"GraViti introduces a graph-level VAE with relaxed permutation invariance that maps whole graphs to latent vectors, achieves strong reconstruction on large molecular datasets, and generates valid samples by learning constraints directly from graph-level representations.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2605.09452","ref_index":8,"ref_count":1,"confidence":0.88,"is_internal_anchor":false,"paper_title":"Consistency between dynamical modeling and photometrically derived masses of fireballs","primary_cat":"astro-ph.EP","submitted_at":"2026-05-10T10:07:56+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":7.0,"formal_verification":"none","one_line_summary":"A new three-point inverse solution using the α-β model reconstructs meteoroid masses and bulk densities from limited fireball observations, achieving 88% convergence on the EN catalog and producing a continuous density range of 300-4000 kg m^{-3} instead of discrete PE categories.","context_count":1,"top_context_role":"method","top_context_polarity":"use_method","context_text":"andm= 1, the analytical solution of Eqs. (4) and (5) can be expressed as: m(β, µ, v) =e −β 1−v2 1−µ (6) and 4 y(α, β, v) = ln(2α) +β−ln( Ei(β)− Ei(βv 2)),(7) where Ei(x) = Z x −∞ ezdz z is the exponential integral (see Appendix C of Peña-Asensio and Gritse- vich (2025)), where the ballistic coefficientαand the mass-loss parameterβ are defined as: α= cdρslh0Sbeg 2Mbeg sinγ ,(8) β= (1−µ)c hV 2 beg 2cdH ∗ .(9) This approach simplifies fireball flight modeling by reducing all unknowns to two dimensionless parameters,αandβ, which have clear physical mean- ings and can be uniquely derived from observations for each event. In prac- tice, fireballs are classified into comparable outcome groups based on their specificαandβvalues (Gritsevich et al."},{"citing_arxiv_id":"2604.20365","ref_index":16,"ref_count":1,"confidence":0.88,"is_internal_anchor":false,"paper_title":"Benefits of Low-Cost Bio-Inspiration in the Age of Overparametrization","primary_cat":"cs.RO","submitted_at":"2026-04-22T09:02:14+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":3.0,"formal_verification":"none","one_line_summary":"Shallow MLPs and dense CPGs outperform deeper MLPs and Actor-Critic RL in bounded robot control tasks with limited proprioception, with a Parameter Impact metric indicating extra RL parameters yield no performance gain over evolutionary strategies.","context_count":1,"top_context_role":"method","top_context_polarity":"use_method","context_text":"bottom of the core and the ground which would penalize crawling gaits. 4 https://gymnasium.farama.org/environments/mujoco/ant/ Benefits of Low-Cost Bio-Inspiration 7 3.4 Optimization setup The final variable of adjustment investigated in this work is the optimization al- gorithms. On the one hand, we use the Covariance Matrix Adaptation Evolution- ary Strategy (CMA-ES) [15,16], a powerful gradient-free optimization method successfully used in conjunction with this morphological space [9]. Following usual practice with the former version of the framework, we keep all hyperpa- rameters to their default configuration as detailed in the supplementary material [11]. As the method is gradient-free, we used it for both CPGs and MLPs opti-"},{"citing_arxiv_id":"2604.02730","ref_index":13,"ref_count":1,"confidence":0.88,"is_internal_anchor":false,"paper_title":"PhDLspec: physical-prior embedded deep learning method for spectroscopic determination of stellar labels in high-dimensional parameter space","primary_cat":"astro-ph.GA","submitted_at":"2026-04-03T04:44:21+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":6.0,"formal_verification":"none","one_line_summary":"PhDLspec combines differential spectra from physical stellar models with a transformer to derive approximately 30 stellar parameters from low-resolution spectra hundreds of times faster than traditional calculations.","context_count":1,"top_context_role":"method","top_context_polarity":"use_method","context_text":"the algorithm's strategy parameters, including the step sizeσ, mean valuem, and covariance matrix C. Then, generate individuals in the next gener- ation distribution using mutation operations based on these parameters. The updating expression ofσ, m, andCis as follows: σ(g+1) =σ (g) exp( cσ dσ ( ∥p(g+1) σ ∥ E∥N(0, I)∥ −1)) (12) m(g+1) = µX i=1 wix(g+1) i:λ (13) C(g+1) = (1−c µ X wi)C(g) +c µ λX i=1 wiy(g+1) i:λ y(g+1)T i:λ (14) The fitting procedure is formulated as the minimiza- tion of a reducedχ 2 objective function that quantifies the difference between the observed and model spectra: X 2(θ) = 1 n nX i=1 \u0012 fobs,i(λ)−f i(λ;θ) σi(λ) \u00132 , (15) wheref obs(λ) is the observed spectrum,f(λ;θ) is the model prediction,σ(λ) is the observational uncertainty,"},{"citing_arxiv_id":"2512.24497","ref_index":36,"ref_count":1,"confidence":0.88,"is_internal_anchor":false,"paper_title":"What Drives Success in Physical Planning with Joint-Embedding Predictive World Models?","primary_cat":"cs.AI","submitted_at":"2025-12-30T22:50:03+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":5.0,"formal_verification":"none","one_line_summary":"An empirical study of JEPA world models identifies architecture, training objective, and planning choices that yield a model outperforming DINO-WM and V-JEPA-2-AC on navigation and manipulation tasks.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2507.00480","ref_index":68,"ref_count":1,"confidence":0.88,"is_internal_anchor":false,"paper_title":"Posterior Inference in Latent Space for Scalable Constrained Black-box Optimization","primary_cat":"cs.LG","submitted_at":"2025-07-01T06:55:36+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":7.0,"formal_verification":"none","one_line_summary":"Reformulates constrained black-box optimization as posterior inference in latent space of flow-based models amortized by outsourced diffusion models, claiming superior performance on synthetic and real tasks.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2502.08208","ref_index":9,"ref_count":1,"confidence":0.88,"is_internal_anchor":false,"paper_title":"Exploring Exploration in Bayesian Optimization","primary_cat":"cs.LG","submitted_at":"2025-02-12T08:38:37+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":7.0,"formal_verification":"none","one_line_summary":"Introduces observation traveling salesman distance and observation entropy to quantify exploration in Bayesian optimization acquisition functions and links them to empirical performance.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"2304.11468","ref_index":27,"ref_count":1,"confidence":0.88,"is_internal_anchor":false,"paper_title":"Increasing the Scope as You Learn: Adaptive Bayesian Optimization in Nested Subspaces","primary_cat":"cs.LG","submitted_at":"2023-04-22T19:20:39+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":6.0,"formal_verification":"none","one_line_summary":"BAxUS adapts Bayesian optimization over nested random subspaces to achieve better performance than prior high-dimensional methods while providing theoretical guarantees against failure.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null},{"citing_arxiv_id":"1906.09013","ref_index":17,"ref_count":1,"confidence":0.88,"is_internal_anchor":false,"paper_title":"Local Online Motor Babbling: Learning Motor Abundance of A Musculoskeletal Robot Arm","primary_cat":"cs.RO","submitted_at":"2019-06-21T08:56:29+00:00","verdict":"UNVERDICTED","verdict_confidence":"LOW","novelty_score":5.0,"formal_verification":"none","one_line_summary":"A method combining goal babbling with CMA-ES-based local online motor babbling is used to learn inverse kinematics and explore motor abundance on a 10-DoF musculoskeletal robot arm.","context_count":0,"top_context_role":null,"top_context_polarity":null,"context_text":null}],"limit":50,"offset":0}