·
DataBubble
  • Home
  • Models
  • News
  • Compare
  • Boards
  • Pricing
  • About
  • Newsletter
  • Methodology
  • Contact
Latest
What Transfers from Training-Only Priors? Separating Knowledge, Ordinal Targets, and Optimization46m◆SpectrumQuant: Spectrum Alignment for LLM Quantization46m◆A Mixture-Gated Theory of Anisotropic Representation Dynamics in Transformers46m◆Privacy Accounting for Empirical Denoisers46m◆Foundations of the Scaling Law46m◆Lean Finder V2: Semantic Search for Mathlib That Disambiguates User Queries46m◆The Value of System 1 Models in Medicine46m◆Harmful Responses Are Jailbreak Templates: Transferring Autoregressive Jailbreaks to Diffusion Language Models46m◆Stochastic Inverse Problem: stability, regularization and Wasserstein gradient flow46m◆Quantum-aware Transformer model for state classification46m◆Deep Learning Warm Starts for Trajectory Optimization on the International Space Station46m◆BrainRotViT: Transformer-ResNet Hybrid for Explainable Modeling of Brain Aging from 3D sMRI46m◆MedInsightBench: Evaluating Medical Analytics Agents Through Multi-Step Insight Discovery in Multimodal Medical Data46m◆Isotropic Activation Functions Enable Deindividuated Neurons and Adaptive Topologies46m◆Half-Truths Break Similarity-Based Retrieval46m◆How do standard time series forecasting benchmarks translate to finance? A broad study of risk adjusted performance46m◆Temporal Fair Division in Multi-Agent Systems: From Precise Alternation Metrics to Scalable Coordination Proxies46m◆Causal Representation Learning for Generalisable Recommendation46m◆Instance-dependent Stochastic Lipschitz bandit46m◆Driving Video Retrieval for Complex Queries with Structured Grounding46m◆What Transfers from Training-Only Priors? Separating Knowledge, Ordinal Targets, and Optimization46m◆SpectrumQuant: Spectrum Alignment for LLM Quantization46m◆A Mixture-Gated Theory of Anisotropic Representation Dynamics in Transformers46m◆Privacy Accounting for Empirical Denoisers46m◆Foundations of the Scaling Law46m◆Lean Finder V2: Semantic Search for Mathlib That Disambiguates User Queries46m◆The Value of System 1 Models in Medicine46m◆Harmful Responses Are Jailbreak Templates: Transferring Autoregressive Jailbreaks to Diffusion Language Models46m◆Stochastic Inverse Problem: stability, regularization and Wasserstein gradient flow46m◆Quantum-aware Transformer model for state classification46m◆Deep Learning Warm Starts for Trajectory Optimization on the International Space Station46m◆BrainRotViT: Transformer-ResNet Hybrid for Explainable Modeling of Brain Aging from 3D sMRI46m◆MedInsightBench: Evaluating Medical Analytics Agents Through Multi-Step Insight Discovery in Multimodal Medical Data46m◆Isotropic Activation Functions Enable Deindividuated Neurons and Adaptive Topologies46m◆Half-Truths Break Similarity-Based Retrieval46m◆How do standard time series forecasting benchmarks translate to finance? A broad study of risk adjusted performance46m◆Temporal Fair Division in Multi-Agent Systems: From Precise Alternation Metrics to Scalable Coordination Proxies46m◆Causal Representation Learning for Generalisable Recommendation46m◆Instance-dependent Stochastic Lipschitz bandit46m◆Driving Video Retrieval for Complex Queries with Structured Grounding46m◆
News/Characterizing Narrative Content in Web-scale LLM Pretraining Data
arxiv
PublishedJune 19, 2026 at 4:00 AM
—neutral

Characterizing Narrative Content in Web-scale LLM Pretraining Data

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

arXiv:2606.19468v1 Announce Type: new Abstract: The narrative composition of web-scale LLM pretraining corpora remains largely unexplored even though narrative is a fundamental mode of human communication. We present the first fine-grained study of narrative features in Dolma, a 3-trillion-token ope

Models mentioned
01
  • 01facebook logo
    roberta-base
    facebook/roberta-base
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
    NarraBERT
  • 02
    roberta-base
    facebook/roberta-base
  • 03
    Dolma
Source
↗
arxiv
Read original ↗All from arxiv →
Tags
04
#narrative-analysis#pretraining#llm#dataset
Mentioned companies
01
Facebook

No replies yet. Be first.

Mentioned models
03
  • 01
    NarraBERT
  • 02
    roberta-base
    facebook/roberta-base
  • 03
    Dolma
Source
↗
arxiv
Read original ↗All from arxiv →
Tags
04
#narrative-analysis#pretraining#llm#dataset
Mentioned companies
01
Facebook

Related coverage

More from ARXIV
arxivWhat Transfers from Training-Only Priors? Separating Knowledge, Ordinal Targets, and Optimization46marxivSpectrumQuant: Spectrum Alignment for LLM Quantization46marxivA Mixture-Gated Theory of Anisotropic Representation Dynamics in Transformers46marxivPrivacy Accounting for Empirical Denoisers46m
The Bubble Brief
MON · WED · FRI

Read narrative-analysis 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