AI News

EvoSignal: LLM-Guided Evolutionary Design of Modular Traffic Signal Control Programs (arxiv.org)

Online Resource Allocation with an Endogenous Markov State: Fewer LP Solves Earn More (arxiv.org)

COPC: Coupled Off-Policy Correction for Asynchronous LLM Reinforcement Learning (arxiv.org)

Sequential Pretraining Favors Large Models (arxiv.org)

The Identifiability and Observability of Deep Normalized Attention (arxiv.org)

When does a network's training history predict its future learning better than its current state? Evidence from a response probe and a forecasting screen (arxiv.org)

Coding-Agent Benchmarks Should Match Their Users' Task Flows (arxiv.org)

Closed-Form Noise Calibration Against Membership Inference for Random-Allocation DP-SGD (arxiv.org)

Work While They Sleep: Exploiting Evaluation Latency for Fully Bayesian Optimization (arxiv.org)

Slow Beats Fast at the Kesten-Stigum Threshold: Minimax, Fisher-Information and Belief-Propagation Characterizations of the Information-Computation Gap in Sparse Stochastic Block Models (arxiv.org)

Just for FUNS: LLM-Guided Spatio-Temporal Graph Node Generation for Forecasting Unobserved Node States (arxiv.org)

Exact Dynamics and Finite-Sample Trajectory Recovery of Linear Recursive Feature Machines (arxiv.org)

Best Arm Identification for Bandits with Shifting Means (arxiv.org)

Derivative Gaussian Processes on a Two-Direction Budget (arxiv.org)

Safe Meta-Policy Design with Risk Control (arxiv.org)

Dataset Pruning from First Principles: A Label-Free Linear Programming Approach (arxiv.org)

RoBART: Bayesian Additive Regression Trees with Tree-Specific Rotations (arxiv.org)

Broadly Applicable Approximate MCMC for Switching Stochastic Differential Equations Using Uniformization and Time-Conditioned Factorized Neural Likelihood Estimation (arxiv.org)

Universal Local Error and Realized Amplification for the First-Order EDM Predictor (arxiv.org)

Kinetic Langevin Meets Split Gibbs: Accelerated Posterior Sampling for Imaging Inverse Problems with Diffusion Priors (arxiv.org)

Conformal Prediction for Spatially Dependent Data via Sequential Whitening (arxiv.org)

Towards Calibrated Probabilistic Forecasts for Events of Interest via Outcome-Conditional Recalibration (arxiv.org)

Gaussian Equivalence for Multi-Head Self-Attention (arxiv.org)

A Bayesian Mirror Architecture for Emergent Consciousness: Circular Hierarchies, Self-Manifolds, and Hybrid Event-Self Binding (arxiv.org)

Controlling Dependence in Implicit Generative Models via Spread Mutual Information (arxiv.org)

Extreme Binary Classification: Extreme Value Theory for Extreme Constraint on False Negative (arxiv.org)

Possibilistic Radial Transport for Approximate IM Inference (arxiv.org)

Sparsifying Stochasticity, Not Capacity: Partial Stochasticity via Deep Weight Factorization of Prior Scales (arxiv.org)

Scalable Logistic Gaussian Process Density Regression with Kinetic Langevin Sampling (arxiv.org)

Unpaired Canonical Correlation Analysis (arxiv.org)

Reflected Anchored Langevin Algorithms (arxiv.org)

Adjoint-Based Calibration and Optimal Control of Stochastic Multiscale Bioprocess Digital Twins (arxiv.org)

Covariate-dependent Joint Modeling of Multivariate Ordinal Preferences and Its Connections with Comparison Models (arxiv.org)

Learning Transition Kernels of Jump-Diffusion Processes with Conditional Diffusion Models (arxiv.org)

Quadratic Weak-to-Strong Generalization in Random Feature Networks via Random Matrix Theory (arxiv.org)

Certified by Abstention: Distribution-Free Guarantees for Chain-of-Thought Verifiers at Small Calibration Budgets (arxiv.org)

Careful Judge: Safe and Efficient Human-AI Collaborative Decision Making (arxiv.org)

Trust-Region Optimization for Smooth Potential-Interaction Energies in Wasserstein Space (arxiv.org)

The Impact of Likelihood Tempering on the Limiting Predictive Moments of Variational Bayesian Linear Neural Networks (arxiv.org)

Data Reuse in Non-Stationary Learning (arxiv.org)

LASER: Latent Space Adjoint Matching for Support-Constrained Entropy-Regularized Offline RL (arxiv.org)

Identifiability of a dissipative knowledge-dynamics model: exact recovery under designed excitation, degeneration on observational data (arxiv.org)

Graph-monotone entrywise guarantees for MLE and Rank Centrality on general comparison graphs (arxiv.org)

FedRSPO+: A Heterogeneity-aware Algorithm for Decision-focused Federated Learning (arxiv.org)

Relative Wasserstein Spatial Depth for Cluster Number Selection in Distributional Data (arxiv.org)

Lower Bounds for Parallel Diffusion Sampling (arxiv.org)

An Accuracy--Information Tradeoff for Loss-Difference Conditional Mutual Information (arxiv.org)

Sketched Calibration for Conformal Prediction under Covariate Shift (arxiv.org)

Efficient Best-of-N policy evaluation for inference-time alignment (arxiv.org)

The Symbol of the Surrogate: Measuring Numerical Provenance in Neural PDE Solvers (arxiv.org)