·
DataBubble
  • Home
  • Models
  • News
  • Compare
  • Boards
  • Pricing
  • About
  • Newsletter
  • Methodology
  • Contact
Latest
Advancing next-gen AI with materials science innovation30m◆Gritt exits stealth with $34 million for robots to build solar plants—then, everything else1h◆Capacity and Redundancy Trade-offs in Multi-Task Learning7h◆Predictive Training with Latent Imagination for Visual Quadruped Navigation7h◆Where Not to Learn: Prior-Aligned Training with Subset-based Attribution Constraints for Reliable Decision-Making7h◆Did We Actually Fix It? An Independent Adversarial Stress-Test of Post-Point-Adjustment Evaluation Metrics for Time-Series Anomaly Detection7h◆Supervised Reward Inference7h◆PPO-HSC: An Exploratory Reinforcement Learning Framework Based on Wide-Area Policy Coverage Optimization7h◆RECON: Benchmarking Agent Memory for Compositional Reasoning over Long Contexts7h◆Expected Free Energy as Belief-Dependent Utility for rho-POMDPs7h◆Is Progressive Disclosure All You Need for Long-Context Agents?7h◆Verify, Repair, Repeat, or Stop? Robust Stopping for Noisy Verify-Repair Loops in LLM Agents7h◆Hierarchical Wireless Foundation Model for Multi-Task Optimization7h◆Auto Research for Materials: Auditable AI-Scientist Workflows with Held-Out Transfer7h◆It Depends on the Dataset: When a Brain-Encoding Model's Predicted Responses Beat Their Visual Backbone for Video Memorability7h◆A${}^2$BM: Alignment-Aware Bridge Matching for Image-to-Image Translation7h◆DMFNet: Dual-Backbone Multiscale Fusion Network for Urban Scene Classification7h◆Retrieval is Enough: Training-Free Interpretability with a Tool-Using Agent7h◆RealDESED: A Real-World Domestic Sound Event Detection Benchmark7h◆Kernelized Linear Attention: Breaking the Capacity Wall with Symmetric Cones7h◆Advancing next-gen AI with materials science innovation30m◆Gritt exits stealth with $34 million for robots to build solar plants—then, everything else1h◆Capacity and Redundancy Trade-offs in Multi-Task Learning7h◆Predictive Training with Latent Imagination for Visual Quadruped Navigation7h◆Where Not to Learn: Prior-Aligned Training with Subset-based Attribution Constraints for Reliable Decision-Making7h◆Did We Actually Fix It? An Independent Adversarial Stress-Test of Post-Point-Adjustment Evaluation Metrics for Time-Series Anomaly Detection7h◆Supervised Reward Inference7h◆PPO-HSC: An Exploratory Reinforcement Learning Framework Based on Wide-Area Policy Coverage Optimization7h◆RECON: Benchmarking Agent Memory for Compositional Reasoning over Long Contexts7h◆Expected Free Energy as Belief-Dependent Utility for rho-POMDPs7h◆Is Progressive Disclosure All You Need for Long-Context Agents?7h◆Verify, Repair, Repeat, or Stop? Robust Stopping for Noisy Verify-Repair Loops in LLM Agents7h◆Hierarchical Wireless Foundation Model for Multi-Task Optimization7h◆Auto Research for Materials: Auditable AI-Scientist Workflows with Held-Out Transfer7h◆It Depends on the Dataset: When a Brain-Encoding Model's Predicted Responses Beat Their Visual Backbone for Video Memorability7h◆A${}^2$BM: Alignment-Aware Bridge Matching for Image-to-Image Translation7h◆DMFNet: Dual-Backbone Multiscale Fusion Network for Urban Scene Classification7h◆Retrieval is Enough: Training-Free Interpretability with a Tool-Using Agent7h◆RealDESED: A Real-World Domestic Sound Event Detection Benchmark7h◆Kernelized Linear Attention: Breaking the Capacity Wall with Symmetric Cones7h◆
News/The Norm-Separation Delay Law of Grokking: A First-Principles Theory of Delayed Generalization
arxiv
PublishedMay 6, 2026 at 4:00 AM

The Norm-Separation Delay Law of Grokking: A First-Principles Theory of Delayed Generalization

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

arXiv:2603.13331v2 Announce Type: replace Abstract: Grokking -- the sudden generalisation that appears long after a model has perfectly memorised its training data -- has been widely observed but lacks a quantitative theory explaining the length of the delay. We show that grokking is a norm-driven r

Stay posted· Newsletter

A 5-min weekly brief — top movers, price watch, story of the week.

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

Discussion
Source
↗
arxiv
Read original ↗All from arxiv →

No replies yet. Be first.

Source
↗
arxiv
Read original ↗All from arxiv →

Related coverage

More from ARXIV
arxivCapacity and Redundancy Trade-offs in Multi-Task Learning7harxivPredictive Training with Latent Imagination for Visual Quadruped Navigation7harxivWhere Not to Learn: Prior-Aligned Training with Subset-based Attribution Constraints for Reliable Decision-Making7harxivDid We Actually Fix It? An Independent Adversarial Stress-Test of Post-Point-Adjustment Evaluation Metrics for Time-Series Anomaly Detection7h
The Bubble Brief
WEEKLY

Read AI insights every Tuesday — top movers, new releases, story of the week.

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

Originally published on arxiv ↗
HomeModelsNews