惯性聚合 高效追踪和阅读你感兴趣的博客、新闻、科技资讯
阅读原文 在惯性聚合中打开

推荐订阅源

V
Vulnerabilities – Threatpost
大猫的无限游戏
大猫的无限游戏
M
MIT News - Artificial intelligence
IT之家
IT之家
B
Blog
博
博客园 - 【当耐特】
H
Hackread – Cybersecurity News, Data Breaches, AI and More
AI
AI
S
SegmentFault 最新的问题
N
News | PayPal Newsroom
人人都是产品经理
人人都是产品经理
I
InfoQ
GbyAI
GbyAI
WordPress大学
WordPress大学
Hugging Face - Blog
Hugging Face - Blog
D
DataBreaches.Net
Google DeepMind News
Google DeepMind News
L
LINUX DO - 最新话题
爱范儿
爱范儿
博
博客园 - 叶小钗
雷峰网
雷峰网
腾
腾讯CDC
Recent Announcements
Recent Announcements
F
Fortinet All Blogs
U
Unit 42
C
CERT Recently Published Vulnerability Notes
S
Security Archives - TechRepublic
Cyberwarzone
Cyberwarzone
cs.AI updates on arXiv.org
cs.AI updates on arXiv.org
K
KPMG report finds enterprise disconnect between AI and its ROI | CIO
H
Heimdal Security Blog
A
Arctic Wolf
OSCHINA 社区最新新闻
OSCHINA 社区最新新闻
cs.CL updates on arXiv.org
cs.CL updates on arXiv.org
Google DeepMind News
Google DeepMind News
www.infosecurity-magazine.com
www.infosecurity-magazine.com
Google Online Security Blog
Google Online Security Blog
T
The Blog of Author Tim Ferriss
T
Tailwind CSS Blog
美
美团技术团队
N
Netflix TechBlog - Medium
Last Week in AI
Last Week in AI
T
The Exploit Database - CXSecurity.com
Scott Helme
Scott Helme
S
Security @ Cisco Blogs
Apple Machine Learning Research
Apple Machine Learning Research
Y
Y Combinator Blog
小众软件
小众软件
Jina AI
Jina AI
Cyber Security Advisories - MS-ISAC
Cyber Security Advisories - MS-ISAC

stat updates on arXiv.org

HawkesLLM: Semantic Uncertainty Propagation in Agentic Text Simulation When Is Next-Token Prediction Useful? Marginalization, Ergodicity, Mixture Identifiability, Local Sufficiency, RAG, Tools, and Programming Robust OT-Guided Generative Residual Domain Adaptation for Bike-Sharing Demand Prediction under Temporal Domain Shift Any-Dimensional Invariant Universality Instance-Optimal Estimation with Multiple LLM Judges on a Budget Optimal Dimension-Free Sampling for Regularized Classification Entrywise Error Bounds for Spectral Ranking with Semi-Random Adversaries Training-Free Looped Transformers Diffusion-based Denoising Beats Vanilla Score Matching in Parameter Estimation: A Theoretical Explanation Uncertainty-aware classification and triage of structural heart disease using electrocardiography and echocardiography metrics LLM Sparsity Prior for Robust Feature Selection Operationalizing Individual Fairness via Gradient Descent and Bradley-Terry Models Entropy Equivalence Testing Dropout Universality: Scaling Laws and Optimal Scheduling at the Edge-of-Chaos Symbolic Density Estimation for Discrete Distributions Partial Fusion of Neural Networks: Efficient Tradeoffs Between Ensembles and Weight Aggregation The ASE-LSE Disagreement Landscape: An End-to-End Characterisation of Extremes and Structural Drivers A Tale of Two Cities: Pessimism and Opportunism in Offline Dynamic Pricing Understanding and Improving Noisy Embedding Techniques in Instruction Finetuning KAPLAN: Kolmogorov-Arnold Prognostic Learnable Activation Networks for Survival Analysis Anytime Training with Schedule-Free Spectral Optimization Human-Centered Learning Mechanics: A Dynamical Framework for Entropy-Regulated Representation Learning The Matching Principle: A Geometric Theory of Loss Functions for Nuisance-Robust Representation Learning Finite-Particle Convergence Rates for Conservative and Non-Conservative Drifting Models Proxy-Based Approximation of Shapley and Banzhaf Interactions Symmetrization of Loss Functions for Robust Training of Neural Networks in the Presence of Noisy Labels CASCADE Conformal Prediction: Uncertainty-Adaptive Prediction Intervals for Two-Stage Clinical Decision Support The General Theory of Localization Methods Decision-Path Patterns as Tree Reliability Signals: Path-based Adaptive Weighting for Random Forest Classification CT-OT Flow: Estimating Continuous-Time Dynamics from Discrete Temporal Snapshots Learning-to-Defer in Non-Stationary Time Series via Switching State-Space Models Optimal Query Allocation in Extractive QA with LLMs: A Learning-to-Defer Framework with Theoretical Guarantees Adversarial Robustness in One-Stage Learning-to-Defer Learning-to-Defer with Expert-Conditional Advice Variance Reduction for Expectations with Diffusion Teachers Scalable Reinforcement Learning via Adaptive Batch Scaling TASTE: A Designer-Annotated Multi-Dimensional Preference Dataset for AI-Generated Graphic Design A Two-Parameter Weibull Framework for Diagnosing Transformer Weight Distributions Provably Data-driven Lagrangian Relaxation for Mixed Integer Linear Programming Reducing Diffusion Model Memorization with Higher Order Langevin Dynamics Variance-Reduced Manifold Sampling via Polynomial-Maximization Density Estimation Program Evaluation with Remotely Sensed Outcomes Approximate Machine Unlearning through Manifold Representation Forgetting Guided by Self Mode Connectivity Latent Laplace Diffusion for Irregular Multivariate Time Series Multi-task Linear Regression without Eigenvalue Lower Bounds: Adaptivity, Robustness, and Safety Sample efficient inductive matrix completion with noise and inexact side information Feature Learning in Linear-Width Two-Layer Networks: Two vs. One Step of Gradient Descent On Stability and Decomposition of Sample Quantiles under Heavy-Tailed Distributions Shallow ReLU$^s$ Networks in $L^p$-Type and Sobolev Spaces: Approximation and Path-Norm Controlled Generalization Can Adaptive Gradient Methods Converge under Heavy-Tailed Noise? A Case Study of AdaGrad Markov Chain Decoders Overcome the Heavy-Tail Limitations of Lipschitz Generative Models Symmetry-Compatible Principle for Optimizer Design: Embeddings, LM Heads, SwiGLU MLPs, and MoE Routers Reasoning Models Don't Just Think Longer, They Move Differently TabPFN-3: Technical Report BOOST: A Data-Driven Framework for the Automated Joint Selection of Kernel and Acquisition Functions in Bayesian Optimization Yield Curves Dynamics Using Variational Autoencoders Under No-arbitrage ISOMORPH: A Supply Chain Digital Twin for Simulation, Dataset Generation, and Forecasting Benchmarks Coreset-Induced Conditional Velocity Flow Matching Reframing preprocessing selection as model-internal calibration in near-infrared spectroscopy: A large-scale benchmark of operator-adaptive PLS and Ridge models Towards a holistic understanding of Selection Bias for Causal Effect Identification One-Step Generative Modeling via Wasserstein Gradient Flows Adaptive Calibration in Non-Stationary Environments Keeping Score: Efficiency Improvements in Neural Likelihood Surrogate Training via Score-Augmented Loss Functions Online Learning-to-Defer with Varying Experts RISED: A Pre-Deployment Evaluation Framework for High-Stakes AI Decision-Support Systems, with Application to Healthcare Self-Supervised Laplace Approximation for Bayesian Uncertainty Quantification Exact Stiefel Optimization for Probabilistic PLS: Closed-Form Updates, Error Bounds, and Calibrated Uncertainty Federated Language Models Under Bandwidth Budgets: Distillation Rates and Conformal Coverage Real vs. Semi-Simulated: Rethinking Evaluation for Treatment Effect Estimation Modulated learning for private and distributed regression with just a single sample per client device A Refined Generalization Analysis for Extreme Multi-class Supervised Contrastive Representation Learning When Attention Beats Fourier: Multi-Scale Transformers for PDE Solving on Irregular Domains Spherical Flows for Sampling Categorical Data Grokking or Glitching? How Low-Precision Drives Slingshot Loss Spikes Bayesian Rain Field Reconstruction using Commercial Microwave Links and Diffusion Model Priors Order-Agnostic Autoregressive Modelling with Missing Data Neural Stochastic Differential Equations on Compact State Spaces: Theory, Methods, and Application to Suicide Risk Modeling Query-efficient model evaluation using cached responses Self-Attention as Transport: Limits of Symmetric Spectral Diagnostics Jacobian-Velocity Bounds for Deployment Risk Under Covariate Drift Understanding Self-Supervised Learning via Latent Distribution Matching Knowing When to Quit: A Principled Framework for Dynamic Abstention in LLM Reasoning Efficient Preference Poisoning Attack on Offline RLHF Adaptive Querying with AI Persona Priors Optimal Spatio-Temporal Decoupling for Bayesian Conformal Prediction Electricity price forecasting across Norway's five bidding zones in the post-crisis era Adversarial Robustness of NTK Neural Networks Score-Repellent Monte Carlo: Toward Efficient Non-Markovian Sampler with Constant Memory in General State Spaces Inference of Online Newton Methods with Nesterov's Accelerated Sketching ProEval: Proactive Failure Discovery and Efficient Performance Estimation for Generative AI Evaluation Post-Training Augmentation Invariance S2MAM: Semi-supervised Meta Additive Model for Robust Estimation and Variable Selection Beyond Coefficients: Forecast-Necessity Testing for Interpretable Causal Discovery in Nonlinear Time-Series Models Adaptive Learning via Off-Model Training and Importance Sampling for Fully Non-Markovian Optimal Stochastic Control. Complete version Rare Event Analysis via Stochastic Optimal Control Estimating Continuous Treatment Effects with Two-Stage Kernel Ridge Regression Generative Augmented Inference Beyond Augmented-Action Surrogates for Multi-Expert Learning-to-Defer Feature Learning Dynamics in Infinite-Depth Neural Networks Dataset-Driven Channel Masks in Transformers for Multivariate Time Series
A Stationarity-and-Coupling Criterion for Training-Free Time-Lagged Spectral Embeddings of Multivariate Time Series
[Submitted on 11 Jun 2026] · 2026-06-15 · via stat updates on arXiv.org

View PDF HTML (experimental)

Abstract:We study training-free fixed-length descriptors for multivariate time series and ask not merely whether such a descriptor performs well, but when it can be expected to work at all. Our object of study is $D(\tau)$, built from a time-lagged correlation matrix truncated at the Marchenko-Pastur edge so that only signal-bearing eigenvalues survive and classified by cosine similarity to class centroids with zero learned parameters. The central contribution is not the descriptor but a falsifiable applicability criterion for it. Working from a stationary Gaussian VAR(1) model, we argue that $D(\tau)$ separates two classes when the signals are approximately stationary and the class information lives in their cross-channel temporal coupling rather than in marginal per-channel power. We derive, semi-formally, three consequences: a distinguishability condition, why the static ($\tau=0$) covariance collapses to chance, and why a stationary but power-discriminated paradigm defeats the descriptor. The criterion is operational: a two-part pre-flight test -- an augmented Dickey-Fuller stationarity check and a power-baseline saturation check -- predicts applicability before any training. We validate both halves on a mixed assortment. On four paradigms that satisfy the criterion (Sleep-EDF, BCI-IV-2a, MIT-BIH, ESC-50) the descriptor is competitive with strong baselines at a fraction of their cost, reaching $88.5\pm4.5\%$ under 20-subject leave-one-subject-out on Sleep-EDF on a single CPU thread. On three that violate it -- non-stationary ERPs, and financial-volatility and wearable-stress regimes that are power-discriminated -- it fails exactly as the pre-flight predicts, and these negatives are the more informative half. We are explicit that $D(\tau)$ is not the most accurate representation; its value is a compact, training-free embedding whose domain of validity is known in advance.

Submission history

From: Siddharth Pal [view email]
[v1] Thu, 11 Jun 2026 18:54:21 UTC (120 KB)

此内容由惯性聚合(RSS阅读器)自动聚合整理，仅供阅读参考。 原文来自 — 版权归原作者所有。