·
DataBubble
  • Home
  • Models
  • News
  • Compare
  • Boards
  • Pricing
  • About
  • Newsletter
  • Methodology
  • Contact
Latest
AdaPRA: Process Reward Agents with Adaptive Rubrics for Data Science3h◆Scoped Semantic Invariance as a Foundation for LLM Reliability3h◆GEMS-RAG: Graph-grounded Engineering Multimodal Standards Retrieval-Augmented Generation3h◆Estimation of Propagated Risk in LLM-based Multi-Agent Systems3h◆AgentOrch: How Close Can Global Orchestration Come to Step-by-Step Interaction?3h◆How To Build A Modern Diffusion Language Model3h◆Data Optimization for LLMs: A Survey3h◆LegoSR: A Unified Framework for Self-improving Agent-Guided Scientific Discovery3h◆The Future Is Finer: Multivariate Time Series Super-Resolution Forecasting3h◆SkillShift-Bench: Benchmarking Skill Learning under Environment Evolution for Web Agents3h◆FraudSociety: A Generative Agent-Based Society for the Financial-Fraud Economy3h◆CoEvolve: Counterfactual-Guided Co-Evolution of Memory and Collaboration in Multi-Agent Systems3h◆Learning When to Collaborate: Selective Multi-Agent Medical Reasoning via Uncertainty-Aware Routing3h◆Counterfactual Model Testing Under Distribution Shift3h◆Data Governance for Generative and Agentic AI : Technology, Law, Economics, and Security3h◆When Generated Information Returns: Error Persistence in LLM-Mediated Information Ecosystems3h◆Graph-dependent pooling: adaptive granularity in hierarchical graph neural networks3h◆RAQ-MLA: Role-Aware Quantization for Multi-Head Latent Attention Caches3h◆Jointly Modeling Preferences and Imperfect Decision-Making in Large Language Models3h◆SearchJev: A Calibrated System-1 Model for Search Agents3h◆AdaPRA: Process Reward Agents with Adaptive Rubrics for Data Science3h◆Scoped Semantic Invariance as a Foundation for LLM Reliability3h◆GEMS-RAG: Graph-grounded Engineering Multimodal Standards Retrieval-Augmented Generation3h◆Estimation of Propagated Risk in LLM-based Multi-Agent Systems3h◆AgentOrch: How Close Can Global Orchestration Come to Step-by-Step Interaction?3h◆How To Build A Modern Diffusion Language Model3h◆Data Optimization for LLMs: A Survey3h◆LegoSR: A Unified Framework for Self-improving Agent-Guided Scientific Discovery3h◆The Future Is Finer: Multivariate Time Series Super-Resolution Forecasting3h◆SkillShift-Bench: Benchmarking Skill Learning under Environment Evolution for Web Agents3h◆FraudSociety: A Generative Agent-Based Society for the Financial-Fraud Economy3h◆CoEvolve: Counterfactual-Guided Co-Evolution of Memory and Collaboration in Multi-Agent Systems3h◆Learning When to Collaborate: Selective Multi-Agent Medical Reasoning via Uncertainty-Aware Routing3h◆Counterfactual Model Testing Under Distribution Shift3h◆Data Governance for Generative and Agentic AI : Technology, Law, Economics, and Security3h◆When Generated Information Returns: Error Persistence in LLM-Mediated Information Ecosystems3h◆Graph-dependent pooling: adaptive granularity in hierarchical graph neural networks3h◆RAQ-MLA: Role-Aware Quantization for Multi-Head Latent Attention Caches3h◆Jointly Modeling Preferences and Imperfect Decision-Making in Large Language Models3h◆SearchJev: A Calibrated System-1 Model for Search Agents3h◆
News/Mechanistic Evidence for Preserved-but-Misaligned Representations in Non-IID FedAvg
arxiv
PublishedJuly 16, 2026 at 4:00 AM
—neutral

Mechanistic Evidence for Preserved-but-Misaligned Representations in Non-IID FedAvg

Source
arxiv.orgfull article ↗
Read on arxiv→
Publisher summary· verbatim

arXiv:2512.23043v3 Announce Type: replace Abstract: Federated Averaging (FedAvg) often degrades under non-IID client data, but it remains unclear whether this degradation reflects the loss of client-learned representations or a failure to use representations that are still present. We study this que

Stay posted· Newsletter

Three short emails a week: Monday, Wednesday and Friday. Top movers, new models and the most-referenced story.

// no spam · unsubscribe one-click · free forever

Your email is used only to send The Bubble Brief. Privacy policy

Discussion
Mentioned models
03
  • 01
    FedAvg
  • 02
    CNN
  • 03
    ResNet
Source
↗
arxiv
Read original ↗All from arxiv →
Tags
04
#federated-learning#non-iid-data#machine-learning#deep-learning

No replies yet. Be first.

Mentioned models
03
  • 01
    FedAvg
  • 02
    CNN
  • 03
    ResNet
Source
↗
arxiv
Read original ↗All from arxiv →
Tags
04
#federated-learning#non-iid-data#machine-learning#deep-learning

Related coverage

More from ARXIV
arxivLiveOption: Evaluating LLM Agents in Structured Option Trading with Nonlinear Payoffs3harxivA Computer Vision Approach to Visual Fraud Detection in Phishing Websites Using YOLOv83harxivThe Value of Memory in Language Agents3harxivHyScale: Scaling Long-Context Training for Learned Sparse Attention3h
The Bubble Brief
MON · WED · FRI

Read federated-learning insights every Monday, Wednesday and Friday — top movers, new models, story of the week.

// no spam · unsubscribe one-click · free forever

Your email is used only to send The Bubble Brief. Privacy policy

Originally published on arxiv ↗
PrivacyTermsRefundsLegal notice|AboutMethodologyContact

© 2026 Databubble

Built by Marouane Gazouzi · Readyo, AI interview coach

HomeModelsNews