A certified gradient-based method for contact-rich manipulation that quantifies smoothing-induced errors via set-valued discrepancies and incorporates them into analytical reachable sets for robust affine feedback policies.
hub Mixed citations
The CMA Evolution Strategy: A Tutorial
Mixed citation behavior. Most common role is background (50%).
57
Pith papers citing it
Background
50% of classified citations
abstract
This tutorial introduces the CMA Evolution Strategy (ES), where CMA stands for Covariance Matrix Adaptation. The CMA-ES is a stochastic, or randomized, method for real-parameter (continuous domain) optimization of non-linear, non-convex functions. We try to motivate and derive the algorithm from intuitive concepts and from requirements of non-linear, non-convex search in continuous domain.
hub tools
citation-role summary
background 4
method 2
citation-polarity summary
representative citing papers
Promptbreeder evolves both task prompts and the mutation prompts that improve them using LLMs, outperforming Chain-of-Thought and Plan-and-Solve on arithmetic and commonsense reasoning benchmarks.
Latent Heuristic Search performs continuous optimization over learned embeddings of heuristics, using normalizing flows and LLM prompting to discover competitive solvers for TSP, CVRP, KSP, and OBP.
A Q/D-space framework supplies sufficient order conditions for explicit Runge-Kutta methods and supports a recursive construction of even-order methods with stage count (p²-2p+8)/4.
EVA-0 is a zeroth-order test-time adaptation method that uses scale-invariant loss, anchor-guided optimization, and symmetric two-sided perturbations to enable inference and adaptation in two forward passes, outperforming prior methods on ImageNet-C with ViT-Base.
EGGROLL applies low-rank evolution strategies to train leaky integrate-and-fire spiking neural networks, reaching 79.21% accuracy on N-MNIST with 2.23 times lower per-generation time than full-rank ES.
EvoPref applies NSGA-II evolutionary optimization with archive-based diversity to populations of LoRA adapters, yielding 18% higher preference coverage and 47% lower collapse than gradient descent baselines while matching alignment quality.
ENMP prunes negative LoRA modules via evolutionary search to boost merging performance to new state-of-the-art levels across language and vision tasks.
Proprioceptive distribution matching adapts simulators for legged robot policies by comparing observation and action distributions, reducing sim-to-real gaps with minimal real data and no external sensing.
A bootstrap strategy for non-unitary CFTs uses statistical stability of OPE data across cross-ratios to optimize spectra, reproducing A-series minimal models and yielding candidate solutions for c>1.
Introduces observation traveling salesman distance and observation entropy to quantify exploration in Bayesian optimization acquisition functions and links them to empirical performance.
CoRP consolidates reward-weighted perturbations into a single model via low-rank structure, improving base LLMs by 8.1 points on average while using one-tenth the budget of prior ensembles and one forward pass.
Deep bibliography expansion in literature search achieves high recall while human citations are found to have only 51% moderate relevance compared to 86-88% for AI methods.
ECo-MoE co-optimizes latent robot genotypes and a gated mixture of control experts to improve evolvability in robot body-controller co-design.
KSOS-BO improves acquisition function optimization in Bayesian optimization by casting it as a kernel sum of squares semidefinite program, outperforming Sobol, DE, and CMA-ES baselines on 10/15 benchmarks with 81% average regret reduction.
Recasts sampling-based nonconvex optimization as smoothed gradient descent to obtain non-asymptotic convergence guarantees and introduces the DIDA annealed algorithm that converges to the global optimum.
Deep Boltzmann Quantum States with natural-gradient optimization and annealing-like training match exact or best-known solutions for large infinite-range Ising spin glasses and solve job shop scheduling instances.
QD-LLM applies neuroevolution to prompt embeddings within a quality-diversity framework, producing 46% higher coverage and 41% higher QD-score than QDAIF on HumanEval, MBPP, and creative writing benchmarks.
RGSE adapts text embeddings at test time via evolutionary search, using cosine similarity rewards from high-confidence visual proposals to improve open-vocabulary object detection under distribution shifts.
Global-MPPI integrates kernel SOS global search with MPPI local refinement and graduated non-convexity smoothing to achieve faster convergence and lower costs on high-dimensional contact-rich manipulation tasks.
Introduces a single-number performance measure, file-based benchmarking, and efficient text-file storage to evaluate and compare stopping criteria for EMO algorithms.
CCV-QAOA is a new complex-valued continuous-variable variant of QAOA that solves real and complex multivariate optimization problems via a variational framework.
A flow-matching model derives manipulation strategies from object affordance, adds an adversarial interaction prior, and uses stability simulation to generate natural, effective human-human co-manipulation motions.
A k-nearest-neighbor approach constructs problem-specific algorithm portfolios that outperform both single solvers and the virtual best solver in fixed-budget black-box optimization.
citing papers explorer
-
Certified Gradient-Based Contact-Rich Manipulation via Smoothing-Error Reachable Tubes
A certified gradient-based method for contact-rich manipulation that quantifies smoothing-induced errors via set-valued discrepancies and incorporates them into analytical reachable sets for robust affine feedback policies.
-
Promptbreeder: Self-Referential Self-Improvement Via Prompt Evolution
Promptbreeder evolves both task prompts and the mutation prompts that improve them using LLMs, outperforming Chain-of-Thought and Plan-and-Solve on arithmetic and commonsense reasoning benchmarks.
-
Latent Heuristic Search: Continuous Optimization for Automated Algorithm Design
Latent Heuristic Search performs continuous optimization over learned embeddings of heuristics, using normalizing flows and LLM prompting to discover competitive solvers for TSP, CVRP, KSP, and OBP.
-
Low Stage High Order Explicit Runge--Kutta Methods via Q- and D-Conditions: General Theory and Efficient Recursive Construction
A Q/D-space framework supplies sufficient order conditions for explicit Runge-Kutta methods and supports a recursive construction of even-order methods with stage count (p²-2p+8)/4.
-
EVA-0: Test-Time Model Evolution with Only Two Forward Passes per Sample
EVA-0 is a zeroth-order test-time adaptation method that uses scale-invariant loss, anchor-guided optimization, and symmetric two-sided perturbations to enable inference and adaptation in two forward passes, outperforming prior methods on ImageNet-C with ViT-Base.
-
Gradient-Free Training of Spiking Neural Networks via Low-Rank Evolution Strategies
EGGROLL applies low-rank evolution strategies to train leaky integrate-and-fire spiking neural networks, reaching 79.21% accuracy on N-MNIST with 2.23 times lower per-generation time than full-rank ES.
-
EvoPref: Multi-Objective Evolutionary Optimization Discovers Diverse LLM Alignments Beyond Gradient Descent
EvoPref applies NSGA-II evolutionary optimization with archive-based diversity to populations of LoRA adapters, yielding 18% higher preference coverage and 47% lower collapse than gradient descent baselines while matching alignment quality.
-
Evolutionary Negative Module Pruning for Better LoRA Merging
ENMP prunes negative LoRA modules via evolutionary search to boost merging performance to new state-of-the-art levels across language and vision tasks.
-
Simulator Adaptation for Sim-to-Real Learning of Legged Locomotion via Proprioceptive Distribution Matching
Proprioceptive distribution matching adapts simulators for legged robot policies by comparing observation and action distributions, reducing sim-to-real gaps with minimal real data and no external sensing.
-
Bootstrapping non-unitary CFTs
A bootstrap strategy for non-unitary CFTs uses statistical stability of OPE data across cross-ratios to optimize spectra, reproducing A-series minimal models and yielding candidate solutions for c>1.
-
Exploring Exploration in Bayesian Optimization
Introduces observation traveling salesman distance and observation entropy to quantify exploration in Bayesian optimization acquisition functions and links them to empirical performance.
-
Consolidating Rewarded Perturbations for LLM Post-Training
CoRP consolidates reward-weighted perturbations into a single model via low-rank structure, improving base LLMs by 8.1 points on average while using one-tenth the budget of prior ensembles and one forward pass.
-
Rethinking Literature Search Evaluation: Deep Research Helps, and Human Citation Lists Are Not a Ground Truth
Deep bibliography expansion in literature search achieves high recall while human citations are found to have only 51% moderate relevance compared to 86-88% for AI methods.
-
ECo-MoE: Embodiment-Conditioned Mixture of Experts Increases the Evolvability of Robots
ECo-MoE co-optimizes latent robot genotypes and a gated mixture of control experts to improve evolvability in robot body-controller co-design.
-
KSOS-BO: Improving Sampling in Bayesian Optimization via Kernel Sum of Squares
KSOS-BO improves acquisition function optimization in Bayesian optimization by casting it as a kernel sum of squares semidefinite program, outperforming Sobol, DE, and CMA-ES baselines on 10/15 benchmarks with 81% average regret reduction.
-
Global Convergence of Sampling-Based Nonconvex Optimization through Diffusion-Style Smoothing
Recasts sampling-based nonconvex optimization as smoothed gradient descent to obtain non-asymptotic convergence guarantees and introduces the DIDA annealed algorithm that converges to the global optimum.
-
Solving Classical and Quantum Spin Glasses with Deep Boltzmann Quantum States
Deep Boltzmann Quantum States with natural-gradient optimization and annealing-like training match exact or best-known solutions for large infinite-range Ising spin glasses and solve job shop scheduling instances.
-
Parameter-Efficient Neuroevolution for Diverse LLM Generation: Quality-Diversity Optimization via Prompt Embedding Evolution
QD-LLM applies neuroevolution to prompt embeddings within a quality-diversity framework, producing 46% higher coverage and 41% higher QD-score than QDAIF on HumanEval, MBPP, and creative writing benchmarks.
-
Reward-Guided Semantic Evolution for Test-time Adaptive Object Detection
RGSE adapts text embeddings at test time via evolutionary search, using cosine similarity rewards from high-confidence visual proposals to improve open-vocabulary object detection under distribution shifts.
-
Global Sampling-Based Trajectory Optimization for Contact-Rich Manipulation via KernelSOS
Global-MPPI integrates kernel SOS global search with MPPI local refinement and graduated non-convexity smoothing to achieve faster convergence and lower costs on high-dimensional contact-rich manipulation tasks.
-
Benchmarking Stopping Criteria for Evolutionary Multi-objective Optimization
Introduces a single-number performance measure, file-based benchmarking, and efficient text-file storage to evaluate and compare stopping criteria for EMO algorithms.
-
A Complex-Valued Continuous-Variable Quantum Approximation Optimization Algorithm (CCV-QAOA)
CCV-QAOA is a new complex-valued continuous-variable variant of QAOA that solves real and complex multivariate optimization problems via a variational framework.
-
Stability-Driven Motion Generation for Object-Guided Human-Human Co-Manipulation
A flow-matching model derives manipulation strategies from object affordance, adds an adversarial interaction prior, and uses stability simulation to generate natural, effective human-human co-manipulation motions.
-
Similarity-based Portfolio Construction for Black-box Optimization
A k-nearest-neighbor approach constructs problem-specific algorithm portfolios that outperform both single solvers and the virtual best solver in fixed-budget black-box optimization.
-
On the Generalization Bounds of Symbolic Regression with Genetic Programming
Derives a generalization bound for GP-based symbolic regression that decomposes the gap into structure-selection complexity and constant-fitting complexity under tree constraints.
-
Optimal Majoranas in Mesoscopic Kitaev Chains
Microscopic treatment of the hybrid segment in mesoscopic Kitaev chains shows that Andreev bound state parity crossings define optimal sweet spots for localized Majoranas with large gaps.
-
Trajectory-based actuator identification via differentiable simulation
Differentiable simulation enables torque-sensor-free actuator model identification from trajectory data, achieving 1.88x better position tracking than a stand-trained baseline and 46% longer travel in downstream locomotion policies.
-
GeoPAS: Geometric Probing for Algorithm Selection in Continuous Black-Box Optimization
GeoPAS uses multi-scale 2D geometric slices of optimization landscapes with validity-mask pooling and a learned-plus-prior composite score to select from 12 solvers, cutting mean relative expected running time from 30.37 to around 3.1-3.6 on within-suite benchmarks.
-
Sumo: Dynamic and Generalizable Whole-Body Loco-Manipulation
Test-time steering of pre-trained whole-body policies via sample-based planning lets legged robots generalize dynamic loco-manipulation to varied heavy objects and tasks without additional training or tuning.
-
PhDLspec: physical-prior embedded deep learning method for spectroscopic determination of stellar labels in high-dimensional parameter space
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.
-
Depth Augmented and FE Free 3D/2D Liver Registration for Laparoscopic Liver AR
A depth-augmented rigid pose refinement combined with a patient-specific statistical deformation model from NICP correspondences achieves 14.73 mm mean TRE for 3D-2D liver registration in controlled laparoscopic settings.
-
Evolution With Purpose: Hierarchy-Informed Optimization of Whole-Brain Models
Hierarchy-informed curricular optimization of heterogeneous whole-brain models enables generalization to new subjects and prediction of behavioral abilities from parameters.
-
Sample-Efficient Optimisation over the Outputs of Generative Models
O3 uses surrogate latent spaces extracted from generative models to perform sample-efficient black-box optimization over their outputs, outperforming direct sampling and original-latent optimization on image and protein tasks.
-
Learning Evolution via Optimization Knowledge Adaptation
OKAEM is a unified learnable evolutionary framework that uses attention-based operators for pre-training on prior knowledge and real-time self-tuning adaptation.
-
Model Merging to Evolution: Parameter Space Exploration for Expert Models
MERGEvolve unifies model merging with evolutionary strategies to explore outside convex parameter space and achieves competitive benchmark performance.
-
Bounding the Effect of HOD Assumptions on Small-Scale Clustering Constraints
The fraction of AbacusSummit cosmologies excluded at 3σ by small-scale clustering multipoles drops from 81% to 25% when moving from fixed HOD parameters to broad marginalization over the five-parameter HOD model.
-
Is the Dark Comet 1998 KY$_{26}$ the Spacecraft Phobos 1?
Orbital calculations indicate that 1998 KY26 could be Phobos 1 if the probe executed two burns totaling 1.9 km/s after its 1988 loss.
-
Dark Quest II: A Wide-Coverage Neural Network Emulator of the Nonlinear Matter Power Spectrum Across Extended Cosmologies
A neural network emulator trained on multi-resolution N-body simulations reproduces the nonlinear matter power spectrum to subpercent accuracy up to the Nyquist scale across an extended nine-dimensional cosmological parameter space.
-
Black-Box Optimization of Mixed Binary-Continuous Variables: Challenges and Opportunities in Evolutionary Model Merging
Data flow space model merging is formalized as a mixed binary-continuous black-box optimization problem, where a structured approach respecting variable dependencies achieves 6.7% higher accuracy and 51.4% smaller search space than unstructured methods on real language models.
-
Distributed Quantum-Enhanced Optimization: A Topographical Preconditioning Approach for High-Dimensional Search
D-QEO framework uses quantum topographical preconditioning on separable functions via small parallel subcircuits to generate seeds that accelerate classical global optimization and avoid exponential failure rates.
-
Rapid LoRA Aggregation for Wireless Channel Adaptation in Open-Set Radio Frequency Fingerprinting
LoRA pretraining per environment plus weighted aggregation at inference cuts EER by 15% and training time by 83% for open-set RFF authentication under varying channels.
-
What Drives Success in Physical Planning with Joint-Embedding Predictive World Models?
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.
-
Emergence of Internal State-Modulated Swarming in Multi-Agent Patch Foraging System
In a simulated multi-agent foraging environment, evolved neural controllers lead to swarming that is modulated by the agents' internal resource levels, with hidden states encoding resource information.
-
Machine Learning in the 2HDM2S model for Dark Matter
A 2HDM extended by two real scalar singlets is scanned with evolutionary strategies to locate regions satisfying vacuum, unitarity, oblique-parameter, collider and dark-matter constraints.
-
Sampling-Based Global Optimal Control and Estimation via Semidefinite Programming
KernelSOS is shown to be competitive on robot localization and to improve solution quality in high-dimensional trajectory optimization when paired with local solvers.
-
Diffusion Models are Evolutionary Algorithms
Diffusion models are evolutionary algorithms via a denoising-evolution equivalence, yielding Diffusion Evolution that outperforms mainstream EAs on multi-optima tasks.
-
CC-VPSTO: Chance-Constrained Via-Point-Based Stochastic Trajectory Optimisation for Online Robot Motion Planning under Uncertainty
CC-VPSTO formulates stochastic trajectory optimization as a chance-constrained problem, approximates it with Monte Carlo sampling and padding, and integrates it into MPC for online robot motion planning under uncertainty.
-
Local Online Motor Babbling: Learning Motor Abundance of A Musculoskeletal Robot Arm
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.
-
Simplifying Flow Matching Transformations with Low-Rank Mixture Models
MPPCA mixtures as latent densities for normalizing flows reduce transformation complexity via better KL alignment, yielding faster convergence and better generation than standard normal baselines.
-
A Deep Dive into Baryon Asymmetry -- the C2HDM
New BSMPT implementation of baryon asymmetry computation using WKB transport equations with moment truncations and VEV profile solving, validated in the C2HDM with uncertainty and GW interplay analysis.