·
DataBubble
  • Home
  • Models
  • News
  • Compare
  • Boards
  • Pricing
  • About
  • Newsletter
  • Methodology
  • Contact
Latest
How Apple’s big lawsuit could disrupt OpenAI’s IPO plans1h◆Apple’s plot to crush OpenAI1h◆Fine-tune video and image models at scale with NVIDIA NeMo Automodel and 🤗 Diffusers3h◆Patreon stops asking AI bots not to scrape — and starts blocking them3h◆Apple’s lawsuit couldn’t come at a worse time for OpenAI5h◆Why the first GPU financiers are turning to inference chips in a $400 million deal7h◆A scorecard for the AI age9h◆The risk of weather data sabotage is rising10h◆The Steering Budget: Examples beat Knobs15h◆Polestar: Drift-Aware Cache Calibration and Token Commitment for Efficient Inference of Diffusion LLMs15h◆RxBrain: Embodied Cognition Foundation Model with Joint Language-Visual Reasoning and Imagination15h◆When a Verified World Model Still Loses: Play-Adequacy vs Prediction-Accuracy in LLM-Synthesized Code World Models15h◆SAGA: Schema-Aware Grounding for Agentic Text-to-SPARQL Generation15h◆ReasFlow: Assisting Reasoning-Centric Scientific Discovery in Applied Mathematics via a Knowledge-Based Multi-Agent System15h◆LBA: Textual Hard-Label Adversarial Attack under Low Query Budgets15h◆Eta Given Delta: Defining LLM Tool Efficiency With Marginal Tool Utility15h◆Simplicity Paradox: Debunking myths about prompting and datasets for LLM evaluation15h◆LIGO-PINN: Learned Initialization via Gated Optimization to Alleviate Convergence Failures in Physics Informed Neural Networks15h◆OmniaBench: Benchmarking General AI Agents Across Diverse Scenarios15h◆When AI Blurs the Boundaries of Contribution: An Empirical Study of Authorship Calibration15h◆How Apple’s big lawsuit could disrupt OpenAI’s IPO plans1h◆Apple’s plot to crush OpenAI1h◆Fine-tune video and image models at scale with NVIDIA NeMo Automodel and 🤗 Diffusers3h◆Patreon stops asking AI bots not to scrape — and starts blocking them3h◆Apple’s lawsuit couldn’t come at a worse time for OpenAI5h◆Why the first GPU financiers are turning to inference chips in a $400 million deal7h◆A scorecard for the AI age9h◆The risk of weather data sabotage is rising10h◆The Steering Budget: Examples beat Knobs15h◆Polestar: Drift-Aware Cache Calibration and Token Commitment for Efficient Inference of Diffusion LLMs15h◆RxBrain: Embodied Cognition Foundation Model with Joint Language-Visual Reasoning and Imagination15h◆When a Verified World Model Still Loses: Play-Adequacy vs Prediction-Accuracy in LLM-Synthesized Code World Models15h◆SAGA: Schema-Aware Grounding for Agentic Text-to-SPARQL Generation15h◆ReasFlow: Assisting Reasoning-Centric Scientific Discovery in Applied Mathematics via a Knowledge-Based Multi-Agent System15h◆LBA: Textual Hard-Label Adversarial Attack under Low Query Budgets15h◆Eta Given Delta: Defining LLM Tool Efficiency With Marginal Tool Utility15h◆Simplicity Paradox: Debunking myths about prompting and datasets for LLM evaluation15h◆LIGO-PINN: Learned Initialization via Gated Optimization to Alleviate Convergence Failures in Physics Informed Neural Networks15h◆OmniaBench: Benchmarking General AI Agents Across Diverse Scenarios15h◆When AI Blurs the Boundaries of Contribution: An Empirical Study of Authorship Calibration15h◆
News/Feature Repulsion and Spectral Lock-in: An Empirical Study of Two-Layer Network Grokking
arxiv
PublishedMay 13, 2026 at 4:00 AM

Feature Repulsion and Spectral Lock-in: An Empirical Study of Two-Layer Network Grokking

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

arXiv:2605.08119v1 Announce Type: cross Abstract: Tian (2025) proves a repulsion theorem (Theorem 6) for the matrix $ B = (\widetilde{F}^\top \widetilde{F} + \eta I)^{-1} $ during the interactive feature-learning stage of grokking: similar features have negative off-diagonal entries $ B_{j\ell} $, p

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
arxivThe Steering Budget: Examples beat Knobs15harxivPolestar: Drift-Aware Cache Calibration and Token Commitment for Efficient Inference of Diffusion LLMs15harxivRxBrain: Embodied Cognition Foundation Model with Joint Language-Visual Reasoning and Imagination15harxivWhen a Verified World Model Still Loses: Play-Adequacy vs Prediction-Accuracy in LLM-Synthesized Code World Models15h
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