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

U
Unit 42
V
V2EX
Martin Fowler
Martin Fowler
博
博客园 - Franky
P
Proofpoint News Feed
P
Palo Alto Networks Blog
H
Hackread – Cybersecurity News, Data Breaches, AI and More
B
Blog
The Register - Security
The Register - Security
Latest news
Latest news
S
Security @ Cisco Blogs
Simon Willison's Weblog
Simon Willison's Weblog
Recorded Future
Recorded Future
大猫的无限游戏
大猫的无限游戏
M
Microsoft Research Blog - Microsoft Research
Scott Helme
Scott Helme
T
Tailwind CSS Blog
cs.CV updates on arXiv.org
cs.CV updates on arXiv.org
Cyber Security Advisories - MS-ISAC
Cyber Security Advisories - MS-ISAC
Application and Cybersecurity Blog
Application and Cybersecurity Blog
T
True Tiger Recordings
有赞技术团队
有赞技术团队
I
Intezer
Cisco Talos Blog
Cisco Talos Blog
Hacker News - Newest:
Hacker News - Newest: "LLM"
The GitHub Blog
The GitHub Blog
cs.AI updates on arXiv.org
cs.AI updates on arXiv.org
T
Tenable Blog
博
博客园 - 叶小钗
Hugging Face - Blog
Hugging Face - Blog
Hacker News: Ask HN
Hacker News: Ask HN
S
Security Archives - TechRepublic
F
Future of Privacy Forum
爱范儿
爱范儿
PCI Perspectives
PCI Perspectives
H
Help Net Security
让小产品的独立变现更简单 - ezindie.com
让小产品的独立变现更简单 - ezindie.com
T
The Blog of Author Tim Ferriss
MyScale Blog
MyScale Blog
N
Netflix TechBlog - Medium
罗
罗磊的独立博客
Apple Machine Learning Research
Apple Machine Learning Research
MongoDB | Blog
MongoDB | Blog
Security Latest
Security Latest
美
美团技术团队
博
博客园 - 三生石上(FineUI控件)
S
Schneier on Security
量
量子位
C
CERT Recently Published Vulnerability Notes
SecWiki News
SecWiki News

cs.LG updates on arXiv.org

TBP-mHC: full expressivity for manifold-constrained hyper connections through transportation polytopes Can Transformers Learn to Verify During Backtracking Search? Same Architecture, Different Capacity: Optimizer-Induced Spectral Scaling Laws SCI-Defense: Defending Manipulation Attacks from Generative Engine Optimization PEARL: Unbiased Percentile Estimation via Contrastive Learning for Industrial-Scale Livestream Recommendation Double descent for least-squares interpolation on contaminated data: A simulation study Quantitative coronary calcification analysis for prediction of myocardial ischemia using non-contrast CT calcium scoring Objective-Induced Bias and Search Dynamics in Multiobjective Unsupervised Feature Selection Learning Causal Orderings for In-Context Tabular Prediction Detecting Atypical Clients in Federated Learning via Representation-Level Divergence DualOptim+: Bridging Shared and Decoupled Optimizer States for Better Machine Unlearning in Large Language Models On the Sample Complexity of Discounted Reinforcement Learning with Optimized Certainty Equivalents One LR Doesn't Fit All: Heavy-Tail Guided Layerwise Learning Rates for LLMs Optimal Guarantees for Auditing Rényi Differentially Private Machine Learning Decomposing Ensemble Spread in Lorenz '96 With Learned Stochastic Parameterizations LABO: LLM-Accelerated Bayesian Optimization through Broad Exploration and Selective Experimentation Alike Parts: A Feature-Informed Approach to Local and Global Prototype Explanations Long-term Fairness with Selective Labels An Improved Adaptive PID Optimizer with Enhanced Convergence and Stability for Deep Learning A Reproducible Log-Driven AutoML Framework for Interpretable Pipeline Optimization in Healthcare Risk Prediction Expectation Consistency Loss: Rethink Confidence Calibration under Covariate Shift The Illusion of Reasoning: Exposing Evasive Data Contamination in LLMs via Zero-CoT Truncation Memory-R2: Fair Credit Assignment for Long-Horizon Memory-Augmented LLM Agents Dropout Universality: Scaling Laws and Optimal Scheduling at the Edge-of-Chaos The Attribution Impossibility: No Feature Ranking Is Faithful, Stable, and Complete Under Collinearity What are the Right Symmetries for Formal Theorem Proving? Physics-Informed Generative Solver: Bridging Data-Driven Priors and Conservation Laws for Stable Spatiotemporal Field Reconstruction EmoTrack: Robust Depression Tracking from Counseling Transcripts across Session Regimes Beyond Single Slot: Joint Optimization for Multi-Slot Guaranteed Display Advertising Aerodynamic force reconstruction using physics-informed Gaussian processes Tailoring Teaching to Aptitude: Direction-Adaptive Self-Distillation for LLM Reasoning Short-Term-to-Long-Term Memory Transfer for Knowledge Graphs under Partial Observability Provable Robustness against Backdoor Attacks via the Primal-Dual Perspective on Differential Privacy Three Costs of Amortizing Gaussian Process Inference with Neural Processes ChronoMedicalWorld: A Medical World Model for Learning Patient Trajectories from Longitudinal Care Data MMD-Balls as Credal Sets: A PAC-Bayesian Framework for Epistemic Uncertainty in Test-Time Adaptation Representation Gap: Explaining the Unreasonable Effectiveness of Neural Networks from a Geometric Perspective Dynamic Mixture of Latent Memories for Self-Evolving Agents Toward Understanding Adversarial Distillation: Why Robust Teachers Fail No Epoch Like the Present: Robust Climate Emulation Requires Out-of-Distribution Generalisation Reasoning through Verifiable Forecast Actions: Consistency-Grounded RL for Financial LLMs Leveraging Self-Paced Curriculum Learning for Enhanced Modality Balance in Multimodal Conversational Emotion Recognition A Boundary-Layer Mechanism for One-Third Scaling in Online Softmax Classification Temporal Contrastive Transformer for Financial Crime Detection: Self-Supervised Sequence Embeddings via Predictive Contrastive Coding On-Policy Consistency Training Improves LLM Safety with Minimal Capability Degradation IKNO: Infinite-order Kernel Neural Operators Holomorphic Neural ODEs with Kolmogorov-Arnold Networks for Interpretable Discovery of Complex Dynamics Uniform Diffusion Models Revisited: Leave-One-Out Denoiser and Absorbing State Reformulation One-Way Policy Optimization for Self-Evolving LLMs Explainable AI for Data-Driven Design of High-Dimensional Predictive Studies Towards Explainability of SLMs by investigating Token Level Activation TONIC: Token-Centric Semantic Communication for Task-Oriented Wireless Systems Reinforced Graph of Thoughts: RL-Driven Adaptive Prompting for LLMs Embedding-Based Federated Learning with Runtime Governance for Iron Deficiency Prediction Models Can Model, But Can't Bind: Structured Grounding in Text-to-Optimization ARC-STAR: Auditable Post-Hoc Correction for PDE Foundation Models $\textit{BlockFormer}$ : Transformer-based inference from interaction maps I-SAFE: Wasserstein Coherence Metrics for Structural Auditing of Scientific AI Models Manifold-Guided Attention Steering Adaptive Measurement Allocation for Learning Kernelized SVMs Under Noisy Observations PeakFocus: Bridging Peak Localization and Intensity Regression via a Unified Multi-Scale Framework for Electricity Load Forecasting Equilibrium Propagation and Hamiltonian Inference in the Diffusive Fitzhugh-Nagumo Model Correcting Class Imbalance in Prior-Data Fitted Networks for Tabular Classification CausalGuard: Conformal Inference under Graph Uncertainty Machine learning prediction of obstructive coronary artery disease using opportunistic coronary calcium and epicardial fat assessments from CT calcium scoring scans Beyond Scalar Objectives: Expert-Feedback-Driven Autonomous Experimentation for Scientific Discovery at the Nanoscale Position: The Time for Sampling Is Now! Charting a New Course for Bayesian Deep Learning From Sequential Nodes to GPU Batches: Parallel Branch and Bound for Optimal $k$-Sparse GLMs SepsisAI Orchestrator: A Containerized and Scalable Platform for Deploying AI Models and Real-Time Monitoring in Early Sepsis Detection Symbolic Density Estimation for Discrete Distributions How Sparsity Allocation Shapes Label-Free Post-Pruning Recoverability Prototype-Guided Classification Sub-Task Decoupling Framework: Enhancing Generalization and Interpretability for Multivariate Time Series From Snapshots to Trajectories: Learning Single-Cell Gene Expression Dynamics via Conditional Flow Matching Can Breath Biomarkers Causally Influence Blood Glucose? Investigating VOC-Mediated Modulation in Diabetes VeriScale: Adversarial Test-Suite Scaling for Verifiable Code Generation When to Switch, Not Just What: Transition Quality Prediction in Clash Royale Beyond Euclidean Proximity: Repairing Latent World Models with Horizon-Matched Trajectory Reachability Metrics How Many Different Outputs Can a Transformer Generate? OPPO: Bayesian Value Recursion for Token-Level Credit Assignment in LLM Reasoning Don't Forget the Critic: Value-Based Data Rehearsal for Multi-Cyclic Continual Reinforcement Learning Target-Aligned Bellman Backup for Cross-domain Offline Reinforcement Learning AutoMCU: Feasibility-First MCU Neural Network Customization via LLM-based Multi-Agent Systems Partial Fusion of Neural Networks: Efficient Tradeoffs Between Ensembles and Weight Aggregation AgForce Enables Antigen-conditioned Generative Antibody Design Tabular foundation models for robust calibration of near-infrared chemical sensing data Discovering Entity-Conditioned Lag Heterogeneity: A Lag-Gated Neural Audit Framework for Panel Time Series Harnesses for Inference-Time Alignment over Execution Trajectories Provable Joint Decontamination for Benchmarking Multiple Large Language Models Algebraic Machine Learning for Small-to-Medium Datasets Is Competitive against Strong Standard Baselines Chebyshev Policies and the Mountain Car Problem: Reinforcement Learning for Low-Dimensional Control Tasks CASE-NET: Deep Spatio-Temporal Representation Learning via Causal Attention and Channel Recalibration for Multivariate Time Series Classification Predicting Performance of Symbolic and Prompt Programs with Examples Riemannian geometry meets fMRI: the advantages of modeling correlation manifolds and eigenvector subspaces Calibration, Uncertainty Communication, and Deployment Readiness in CKD Risk Prediction: A Framework Evaluation Study Bandit Convex Optimization with Gradient Prediction Adaptivity stable-worldmodel: A Platform for Reproducible World Modeling Research and Evaluation Noise Schedule Design for Diffusion Models: An Optimal Control Perspective Ex-GraphRAG: Interpretable Evidence Routing for Graph-Augmented LLMs When Are Teacher Tokens Reliable? Position-Weighted On-Policy Self-Distillation for Reasoning ASAP: Attention Sink Anchored Pruning
F-TIS: Harnessing Diverse Models in Collaborative GRPO
Nikolay Blag · 2026-05-23 · via cs.LG updates on arXiv.org

View PDF HTML (experimental)

Abstract:Reinforcement learning methods such as GRPO have seen great popularity in LLM post-training. In GRPO, models produce completions to a set of prompts, which are rewarded, and the policy is updated towards the relatively high reward completions. Due to the auto-regressive nature of models, the generation phase of such style of training can be extremely time consuming. As a solution, prior work has sought to distribute the inference step across many nodes, working parallel. These works assume primarily homogeneous models in the training in order to keep samples as close to on-policy as possible. This assumption may be impractical in decentralized systems, where parties with various computes and preferences may wish to collaborate on the same task. Thus, decentralized training requires an approach that can handle heterogeneous models - different models collaborating on the same tasks. However, this leads to highly off-policy samples presented during training, which prior work has identified that off-policy samples can hurt GRPO convergence. To enable heterogeneity, we propose Filtered Truncated Importance Sampling (F-TIS) - a GRPO-style training paradigm that can use off-policy samples to improve local model's learning. Our framework allows various models to collaborate in the same RL training run while being communication efficient. We extensively evaluate F-TIS in various heterogeneous setups and we show that it exhibits identical final model convergence to purely on-sample training. Furthermore, we observe in some setups better generalization on out-of-distribution tasks than on-policy training, increasing model's performance by up to 12\%.
Comments: Accepted to ICML 2026 Workshop Scalable Learning and Optimization for Efficient Multimodal AI Agents (SCALE)
Subjects: Machine Learning (cs.LG)
Cite as: arXiv:2605.22537 [cs.LG]
  (or arXiv:2605.22537v1 [cs.LG] for this version)
  https://doi.org/10.48550/arXiv.2605.22537

arXiv-issued DOI via DataCite (pending registration)

Submission history

From: Oguzhan Ersoy [view email]
[v1] Thu, 21 May 2026 14:25:24 UTC (290 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。