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

Schneier on Security
Schneier on Security
D
Darknet – Hacking Tools, Hacker News & Cyber Security
T
Tenable Blog
P
Proofpoint News Feed
V
Vulnerabilities – Threatpost
Project Zero
Project Zero
Latest news
Latest news
S
Schneier on Security
C
Cyber Attacks, Cyber Crime and Cyber Security
T
Threatpost
Simon Willison's Weblog
Simon Willison's Weblog
Cyberwarzone
Cyberwarzone
T
The Exploit Database - CXSecurity.com
C
CERT Recently Published Vulnerability Notes
Spread Privacy
Spread Privacy
Cisco Talos Blog
Cisco Talos Blog
T
Troy Hunt's Blog
S
Secure Thoughts
C
Cisco Blogs
Application and Cybersecurity Blog
Application and Cybersecurity Blog
V2EX - 技术
V2EX - 技术
Hacker News: Ask HN
Hacker News: Ask HN
O
OpenAI News
L
LINUX DO - 最新话题
T
Threat Research - Cisco Blogs
Recent Commits to openclaw:main
Recent Commits to openclaw:main
P
Palo Alto Networks Blog
Exploit-DB.com RSS Feed
Exploit-DB.com RSS Feed
博
博客园_首页
月光博客
月光博客
博
博客园 - 【当耐特】
雷峰网
雷峰网
C
CXSECURITY Database RSS Feed - CXSecurity.com
博
博客园 - 叶小钗
aimingoo的专栏
aimingoo的专栏
L
Lohrmann on Cybersecurity
D
DataBreaches.Net
美
美团技术团队
B
Blog
K
KPMG report finds enterprise disconnect between AI and its ROI | CIO
有赞技术团队
有赞技术团队
D
Docker
Jina AI
Jina AI
The GitHub Blog
The GitHub Blog
H
Hacker News: Front Page
cs.CL updates on arXiv.org
cs.CL updates on arXiv.org
AI
AI
Martin Fowler
Martin Fowler
Attack and Defense Labs
Attack and Defense Labs
小众软件
小众软件

cs.LG updates on arXiv.org

Memory-Guided Trust-Region Bayesian Optimization (MG-TuRBO) for High Dimensions EngageTriBoost: Predictive Modeling of User Engagement in Digital Mental Health Intervention Using Explainable Machine Learning Reservoir observer enhanced with residual calibration and attention mechanism Efficient RL Training for LLMs with Experience Replay Wireless Communication Enhanced Value Decomposition for Multi-Agent Reinforcement Learning Adversarial Sensor Errors for Safe and Robust Wind Turbine Fleet Control IKKA: Inversion Classification via Critical Anomalies for Robust Visual Servoing Adaptive Simulation Experiment for LLM Policy Optimization EvoLen: Evolution-Guided Tokenization for DNA Language Model Smartwatch-Based Sitting Time Estimation in Real-World Office Settings Structural Evaluation Metrics for SVG Generation via Leave-One-Out Analysis Loom: A Scalable Analytical Neural Computer Architecture Spectral Geometry of LoRA Adapters Encodes Training Objective and Predicts Harmful Compliance Finite-Sample Analysis of Nonlinear Independent Component Analysis:Sample Complexity and Identifiability Bounds How does Chain of Thought decompose complex tasks? Uncertainty-Aware Transformers: Conformal Prediction for Language Models Adaptive Candidate Point Thompson Sampling for High-Dimensional Bayesian Optimization Using Synthetic Data for Machine Learning-based Childhood Vaccination Prediction in Narok, Kenya Delve into the Applicability of Advanced Optimizers for Multi-Task Learning Bridging SFT and RL: Dynamic Policy Optimization for Robust Reasoning Multi-Agent Decision-Focused Learning via Value-Aware Sequential Communication Predictive Entropy Links Calibration and Paraphrase Sensitivity in Medical Vision-Language Models Efficient Hierarchical Implicit Flow Q-learning for Offline Goal-conditioned Reinforcement Learning Modality-Aware Zero-Shot Pruning and Sparse Attention for Efficient Multimodal Edge Inference The nextAI Solution to the NeurIPS 2023 LLM Efficiency Challenge Feature-Label Modal Alignment for Robust Partial Multi-Label Learning Integrated electro-optic attention nonlinearities for transformers Toward World Models for Epidemiology Tracing the Chain: Deep Learning for Stepping-Stone Intrusion Detection Batch Distillation Data for Developing Machine Learning Anomaly Detection Methods Predicting Metabolic Dysfunction-Associated Steatotic Liver Disease using Machine Learning Methods: A Retrospective Cohort Study Adaptive Tuning of Parameterized Traffic Controllers via Multi-Agent Reinforcement Learning Bandwidth-constrained Variational Message Encoding for Cooperative Multi-agent Reinforcement Learning Neural Two-Stage Stochastic Optimization for Solving Unit Commitment Problem SafeAdapt: Provably Safe Policy Updates in Deep Reinforcement Learning Continuous Orthogonal Mode Decomposition: Haptic Signal Prediction in Tactile Internet Rays as Pixels: Learning A Joint Distribution of Videos and Camera Trajectories PhysInOne: Visual Physics Learning and Reasoning in One Suite Beyond Augmented-Action Surrogates for Multi-Expert Learning-to-Defer FIRE-CIR: Fine-grained Reasoning for Composed Fashion Image Retrieval Detecting Diffusion-generated Images via Dynamic Assembly Forests PDE-regularized Dynamics-informed Diffusion with Uncertainty-aware Filtering for Long-Horizon Dynamics Leave My Images Alone: Preventing Multi-Modal Large Language Models from Analyzing Images via Visual Prompt Injection Regime-Conditional Retrieval: Theory and a Transferable Router for Two-Hop QA Identification and Anonymization of Named Entities in Unstructured Information Sources for Use in Social Engineering Detection Hypergraph Neural Networks Accelerate MUS Enumeration ASTRA: Adaptive Semantic Tree Reasoning Architecture for Complex Table Question Answering Neighbourhood Transformer: Switchable Attention for Monophily-Aware Graph Learning WOMBET: World Model-Based Experience Transfer for Robust and Sample-efficient Reinforcement Learning Low-Data Supervised Adaptation Outperforms Prompting for Cloud Segmentation Under Domain Shift Revisiting the Capacity Gap in Chain-of-Thought Distillation from a Practical Perspective A Mathematical Framework for Temporal Modeling and Counterfactual Policy Simulation of Student Dropout Temporal Dropout Risk in Learning Analytics: A Harmonized Survival Benchmark Across Dynamic and Early-Window Representations MedFormer-UR: Uncertainty-Routed Transformer for Medical Image Classification Dictionary-Aligned Concept Control for Safeguarding Multimodal LLMs Hierarchical Kernel Transformer: Multi-Scale Attention with an Information-Theoretic Approximation Analysis Post-Hoc Guidance for Consistency Models by Joint Flow Distribution Learning SenBen: Sensitive Scene Graphs for Explainable Content Moderation R2G: A Multi-View Circuit Graph Benchmark Suite from RTL to GDSII Policy-Aware Design of Large-Scale Factorial Experiments $p1$: Better Prompt Optimization with Fewer Prompts A Little Rank Goes a Long Way: Random Scaffolds with LoRA Adapters Are All You Need Deep Learning-Based Tracking and Lineage Reconstruction of Ligament Breakup Every Response Counts: Quantifying Uncertainty of LLM-based Multi-Agent Systems through Tensor Decomposition Unified Multimodal Uncertain Inference EfficientSign: An Attention-Enhanced Lightweight Architecture for Indian Sign Language Recognition Skip-Connected Policy Optimization for Implicit Advantage PRAGMA: Revolut Foundation Model 3D-VCD: Hallucination Mitigation in 3D-LLM Embodied Agents through Visual Contrastive Decoding Creator Incentives in Recommender Systems: A Cooperative Game-Theoretic Approach for Stable and Fair Collaboration in Multi-Agent Bandits Evidential Transformation Network: Turning Pretrained Models into Evidential Models for Post-hoc Uncertainty Estimation StructRL: Recovering Dynamic Programming Structure from Learning Dynamics in Distributional Reinforcement Learning Generative 3D Gaussian Splatting for Arbitrary-ResolutionAtmospheric Downscaling and Forecasting From Selection to Scheduling: Federated Geometry-Aware Correction Makes Exemplar Replay Work Better under Continual Dynamic Heterogeneity Needle in a Haystack: One-Class Representation Learning for Detecting Rare Malignant Cells in Computational Cytology Detection of Hate and Threat in Digital Forensics: A Case-Driven Multimodal Approach Semantic Intent Fragmentation: A Single-Shot Compositional Attack on Multi-Agent AI Pipelines Joint Interference Detection and Identification via Adversarial Multi-task Learning HaloProbe: Bayesian Detection and Mitigation of Object Hallucinations in Vision-Language Models R3PM-Net: Real-time, Robust, Real-world Point Matching Network From Dispersion to Attraction: Spectral Dynamics of Hallucination Across Whisper Model Scales AlphaLab: Autonomous Multi-Agent Research Across Optimization Domains with Frontier LLMs Act or Escalate? Evaluating Escalation Behavior in Automation with Language Models Multivariate Time Series Anomaly Detection via Dual-Branch Reconstruction and Autoregressive Flow-based Residual Density Estimation On the Spectral Geometry of Cross-Modal Representations: A Functional Map Diagnostic for Multimodal Alignment Structured Exploration and Exploitation of Label Functions for Automated Data Annotation MolPaQ: Modular Quantum-Classical Patch Learning for Interpretable Molecular Generation CausalVAD: De-confounding End-to-End Autonomous Driving via Causal Intervention Reinforcement-aware Knowledge Distillation for LLM Reasoning SubQuad: Near-Quadratic-Free Structure Inference with Distribution-Balanced Objectives in Adaptive Receptor framework A Horizon-Aware Decision-Support Framework for Demand Forecasting Model Selection in Resilient Production Planning Measurement-Consistent Langevin Corrector for Stabilizing Latent Diffusion Inverse Problem Solvers Multi-agent Adaptive Mechanism Design When & How to Write for Personalized Demand-aware Query Rewriting in Video Search Relational Visual Similarity From Navigation to Refinement: Revealing the Two-Stage Nature of Flow-based Diffusion Models through Oracle Velocity On-the-Fly Adaptation to Quantization: Configuration-Aware LoRA for Efficient Fine-Tuning of Quantized LLMs STCast: Adaptive Boundary Alignment for Global and Regional Weather Forecasting OmniPrism: Learning Disentangled Visual Concept for Image Generation FIT-GNN: Faster Inference Time for GNNs that 'FIT' in Memory Using Coarsening
The Three Regimes of Offline-to-Online Reinforcement Learning
Lu Li, Tianwei Ni, Yihao Sun, Pierre-Luc Bacon · 2025-10-02 · via cs.LG updates on arXiv.org

Offline-to-online reinforcement learning (RL) has emerged as a practical paradigm that leverages offline datasets for pretraining and online interactions for fine-tuning. However, its empirical behavior is highly inconsistent: design choices of online fine-tuning that work well in one setting can fail completely in another. We propose a stability--plasticity principle that can explain this inconsistency: we should preserve the knowledge of pretrained policy or offline dataset during online fine-tuning, whichever is better, while maintaining sufficient plasticity. This perspective identifies three regimes of online fine-tuning, each requiring distinct stability properties. We validate this framework through a large-scale empirical study, finding that the results strongly align with its predictions in 45 of 63 cases, with only 3 opposite mismatches. This work provides a principled framework for guiding design choices in offline-to-online RL based on the relative performance of the offline dataset and the pretrained policy.

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。